Skip to main content

Full text of "Geology of the New York City (Catskill) aqueduct; studies in applied geology covering problems encountered in explorations along the line of the aqued"

See other formats

lEx ICtbrtB 


When you leave, please leave this book 

Because it has been said 
"Ever'thing comes f him who waits 

Except a loaned book." 

Avery Architectural and Fine Arts Library 
Gift of Seymour B. Durst Old York Library 

Digitized by the Internet Archive 
in 2014 

Plate i 


WALOIW ' ' t _, 


vm>-«"i ) i) t. " t4mm v g ..t ^ _____ 

The Catskill and Croton water supply systems of New York city 


Education Department Bulletin 

Published fortnightly by the University of the State of New York 

Entered as second-class matter June 24, 1908, at the Post Office at Albany, H. Y., under 
the act of July 16, 1894 

No. 489 


February 15, 191 1 

New York State Museum 

John M. Clarke, Director 
Museum Bulletin 146 






Introduction and acknowledg- 
ment 5 

I General features 9 

Ch. 1 Catskill water supply 

project 9 

2 Problems encountered in 
the project 17 

3 Relative values of differ- 
ent sources of informa- 
tion and stages of devel- 
opment 25 

4 General geology of the 
region 29 

II Geologic problems of the 
aqueduct 75 

Introduction 75 

Ch. 1 General position of aque- 
duct line 77 

2 Hudson river canyon .... 81 

3 Geological conditions 
affecting the Hudson 
river crossing 97 

4 Geological features in- 
volved in selection of site 

for the Ashokan dam. ... 109 

5 Character and quality of 
the bluestone for struc- 
tural purposes 1J7 


Ch. 6 The Rondout valley sec- 
tion 125 

7 The Wallkffl Valley sec- 
tion 149 

8 Ancient Moodna Valley. . 153 

9 Rock condition of Foundry- 
brook 163 

10 Geology of Sprout brook. 171 

11 Structure of Peekskill 
creek valley 175 

12 Croton lake crossing 183 

13 Geology of the Kensico 
dam site 191 

14 Stone of the Kensico 
quarries 195 

15 The Bryn Mawr siphon.. 201 

16 A study of shaft 13 and 
vicinity on the New Cro- 
ton aqueduct 209 

17 Geological conditions 
affecting the location of 
delivery conduits in New 
York city 215 

18 Areal and structural 
geology south of 59th 
street 231 

19 Special exploration zones. 237 

20 The general question of 
postglacial faulting 271 

Index 277 


191 I 


ru>. 1^ 


Regents of the University 
With years when terms expire 

^ 13 Whitelaw Reid M.A. LL.D. D.C.L. Chancellor New York 

917 St Clair McKelway M.A. LL.D. Vice Chancellor Brooklyn 

919 Daniel Beach Ph.D. LL.D. - - - - - Watkins 

914 Pliny T. Sexton LL.B. LL.D. ----- Palmyra 
912 T. Guilford Smith M.A. C.E. LL.D. - - - Buffalo 

918 William Nottingham M.A. Ph.D. LL.D. - - Syracuse 
922 Chester S. Lord M.A. LL.D. ----- New York 

915 Albert Vander Veer M.D. M.A. Ph.D. LL.D. Albany 
911 Edward Lauterbach M.A. LL.D. - - - - New York 

920 Eugene A. Philbin LL.B. LL.D. - - - - New York 

916 Lucian L. Shedden LL.B. LL.D. - - - - Plattsburg 

921 Francis M. Carpenter ------- Mount Kisco 

Commissioner of Education 

Andrew S. Draper LL.B. LL.D. 

Assistant Commissioners 

Augustus S. Downing M.A. Pd.D. LL.D. First Assistant 
Charles F. Wheelock B.S. LL.D. Second Assistant 
Thomas E. Finegan M.A. Pd.D. Third Assistant 

Director of State Library 

James I. Wyer, Jr, M.L.S. 

Director of Science and State Museum 

John M. Clarke Ph.D. D.Sc. LL.D. 

Chiefs of Divisions 

Administration, George M. Wiley M.A. 
Attendance, James D. Sullivan 

Educational Extension, William R. Eastman MA M.L.S. 

Examinations, Harlan H. Horner B.A. 

Inspections, Frank H. Wood M.A. 

Law, Frank B. Gilbert B.A. 

School Libraries, Charles E: Fitch L.H.D. 

Statistics, Hiram C. Case 

Trades Schools, Arthur D. Dean B.S. 

Visual Instruction, Alfred W. Abrams Ph.B. 

New York Stale Education Department 

Science Division, April 6, 1910 

Hon. Andrew S. Draper LL.D. 

Commissioner of Education 

Sir: The extraordinary engineering operations which have been 
undertaken in the effort to provide the city of New York with an 
adequate waUr supply have illuminated in most unexpected manner 
the geological .structure and history of the region of the Hudson 
valley south of the Catskill mountains. So broad has been the 
scientific scope of this engineering problem and so direct its de- 
pendence on geological structure that the Commissioners of the 
New York City Board of Water Supply early found it of essential 
moment to enlist in their service a corps of trained geologists. 

In 1909 an agreement was effected between the Board of Water 
Supply and the State Geologist, in pursuance of which the geolog- 
ical data acquired in the preliminary and final surveys for the aque- 
duct were intrusted to Dr Charles P. Berkey, a member of the staff 
of the board as well as of the geological survey, for summation and 
presentation of their hroader and more important bearings. 

I transmit to you herewith Dr Berkey's report thereupon, entitled 
Geology of the New York City (Catskill) Aqueduct. It is a 
document of high value not only in enlarging and perfecting our 
knowledge of the geological structure of the commercial center of 
the United States, but its data and conclusions must prove of pro- 
found importance to all large engineering and architectural propo- 
sitions concerned with the region of the lower Hudson valley. 




I therefore submit this, subject to your approval, for immediate 
publication as a bulletin of the State Museum. 

Very respectfully 

John M. Clarke 


State of New York 
Education Department 

commissioner's room 

Approved for publication this Jth day of April 1910 

Education Department Bulletin 

Published fortnightly by the University of the State of New York 

Entered as second-class matter June 24, 1908, at the Post Office at Albany, N. Y., 
under the act of July 16, 1894 

No. 489 ALBANY, N. Y. February 15, 19 1 1 

New York State Museum 

John M. Clarke, Director 
Museum BuUetin 146 







It is the writer's hope that the series of studies brought together 
in this bulletin may help to effect a wider appreciation of the prac- 
tical usefulness of geology. The volume contains a summary of 
the local geologic facts and the general principles found helpful in 
solving some of the problems encountered in a single great engineer- 
ing enterprise. The summary is accompanied by brief discussions 
of the methods employed and of the final results or conclusions 
reached. It is therefore essentially a study in applied geology. 

Seldom has so favorable an opportunity been afforded to follow 
extensive exploratory work and check geologic hypothesis or theory 
by subsequent proof. And still more seldom have engineers in 
charge of similar works so fully appreciated the value of geologic 
investigations and the extent to which they can be utilized as a 

More credit is due to Mr J. Waldo Smith, chief engineer of the 
Board of Water Supply of the City of New York, than to any one 
else for appreciating the importance of the geologic complexity of 



the Catskill Aqueduct problem. His exceptional insight into its 
nature led to the adoption of measures in this direction that are 
now proved to have been fully justified. A staff of geologists has 
been maintained. From time to time engineers of the regular staff 
who have shown unusual aptitude in such investigations have been 
assigned to special duty on geologic exploratory work. In the pre- 
liminary investigations of the Northern Aqueduct, Division Engineer 
James F. Sanborn was very intimately connected with the geologic 
work. With him the writer worked out many field studies that 
later formed the basis of advisory reports, covering locations, kinds 
of explorations to be made, and interpretations of data. No one 
has had a better grasp of both the geologic and the engineering 
aspects than Mr Sanborn. It is with great pleasure that the writer 
acknowledges many valuable suggestions and much help through 
association with him. In the later exploratory work within the city 
similar service has been rendered by Mr John R. Healey, who has 
much to do with the geologic detail of the delivery conduit data. 
The consulting geologists employed by the board were Professors 
James F. Kemp, W. O. Crosby and the writer. 

A special debt is acknowledged to Prof. James F. Kemp, consult- 
ing geologist of the board, whose confidence in the writer's work 
originally brought him into touch with these investigations as an 
assistant, and with whom since that time many joint reports to the 
board have been written. 

Valuable advice and assistance in arranging for the issue of this 
report has been given by Department Engineer Alfred D. Flinn of 
Headquarters Department. For some of the corrections and sug- 
gestions special acknowledgment is made to Department Engineer 
Thaddeus Merrimar . 

The department engineers, Robert Ridgway of the Northern 
Aqueduct, Carlton E. Davis of the Reservoir, Merritt H. Smith, for- 
merly of the Southern Aqueduct, Frank E. Winsor of the Southern 
Aqueduct, William W. Brush and Walter E. Spear of the City De- 
livery have given every facility for gathering geologic data within 
their territory and have contributed largely to the better understand- 
ing of their special fields. 

The geologic matter relating to special problems has been worked 
out with the aid of the division engineers in direct charge in the 
field. Among these must be mentioned L. White of the Esopua 
division, William E. Swift of the Hudson river division, A. A. 
Sproul of the Peekskill division, Lawrence C. Brink of the Wall- 



kill division, J. S. Langthorn of the Ashokan reservoir, Wilson 
Fitch Smith of the Kensico division, T. C. Atwood of the New 
York city delivery division. 

The data included in the tabulation of this bulletin have been 
gathered largely by others. Many of the explanations and conclu- 
sions are the outgrowth of the work of engineer and geologist, 
together. A large number of associates are engaged on this public 
work in such relations to one another that the individuality of each 
is obscured in the common effort to reach an enviable efficiency and 
success for the whole enterprise. 

The combined efforts of many, unselfishly given, have thus 
brought together a total far in excess of what any one individual 
could accomplish. Acknowledgments "should therefore be made to 
those members of the staff of the Board of Water Supply who can 
not in the nature of the case be mentioned by name. Were it not 
for their cooperation the great mass of data here summarized could 
not have been compiled. 

Charles P. Berkey 
Special Geologist, New York State Geological Survey; 
Consulting Geologist New York City Board of Water Supply 
Columbia University, New York City November i, iqio 




New York city obtains its chief water supply from the Croton 
river watershed. Other sources 1 now drawn upon are less important 
although some of them, such as the Long Island underground 
supply, are capable of considerable additional development. The 
average daily consumption of Croton water was approximately 
324,ooo,ooo 2 gallons for 1907. At the present rate of increase of 
population the consequent daily increase in consumption of water 
is 15,000,000 gallons in each succeeding year. 

The entire daily flow of water in the Croton river for the 18 
years from 1879 to l %97 averaged only 348,000,000 gallons. About 
10,000,000 gallons per day is lost by evaporation and seepage 
from existing reservoirs. The records for 40 years, from 1868 to 
1907 make a somewhat better showing. Making no allowance for 
evaporation the average flow amounts to 402,000,000 gallons. With 
due allowance for evaporation, 3 however, this only increases the 
daily supply as now planned by about 47,000,000 gallons. That is, 
the possible total additional water within the Croton watershed 
would suffice for only three years' growth of the city. Much of 
this additional water belongs to periods of excessive precipitation. 
To save it would require additional storage facilities for 3,05,000,- 
000,000 gallons, and, it is estimated, would probably cost $150,- 

1 Brooklyn is in part supplied by these additional sources which furnished 
145,000,000 gallons daily in 1007. 

2 The figures used here as to consumption and capacity and available 
supply are taken from the printed statements of the commissioners of the 
New York City Board of Water Supply in a circular dated April 16, 1908, 
and are based upon the investigation and reports of the corps of engineers 
headed by J. Waldo Smith, chief engineer, John R. Freeman and William 
H. Burr, consulting engineers. The reports of this commission and 
various others that have had the responsibility of investigating the future 
supplies for New York city have been drawn upon freely for such data. 

3 The average rainfall for the past 40 years is about 49 inches per year. 
Only about 48 per cent of this runs into the streams. The rest evap- 
orates or is absorbed by the vegetation or joins underground supplies 
that do not again appear at the surface in the district. 




Taking into account the small relief possible in this direction and 
the certainty that in less than five years the demands of the city 
will be greater than the total capacity of the Croton watershed, it 
is clear that some other source of large and permanent supply is an 
absolute necessity. 

In the search for such additional sources, there has been much 
careful work done by able commissioners. 1 In the meantime, resi- 
dents of certain districts where there are possible supplies have 
taken steps by legislative action to effectually- prevent New York 
city encroaching upon their territory. Criticisms 3 of all kinds 
largely by those only partially informed as to the magnitude and 
complexity of the problem and partly by those ignorant of the 
simplest factors in its solution, have been kept perpetually before 
the public. One needs only a slight acquaintance with such public 
works to realize that it is much easier and more common to criticize 
and raise the cry of corruption or incompetence than it is to give 
really valuable advice or solve a real problem or carry an enterprise 
i «f the most vital public importance to a successful issue. 

It is sufficient here to observe that exhaustive studies of the whole 
question of water supply by competent men have resulted in a 
practically unanimous conclusion that the streams of the Catskill 
mountains are the most satisfactory, economical, reliable, abundant 
and available future source of water. 

1 The Report of John R. Freeman C. E., 1899-1900; Report of the Burr- 
Herring-Freeman Commission, 1902-4; the Studies of the Department of 
Water Supply, Gas and Electricity, 1902-4 ; Investigations of the Board of 
Water Supply, 1905 to the present time. 

2 Acts of the Legislature of 1903-4. 

3 The commonest suggestions neglect the question of permanence or 
constancy of supply. The following sources are often mentioned, (a) Lake 
George, forgetting that this beautiful lake has an abnormally small water- 
shed and could never figure as a large permanent s upply ; (b) artesian 
wells, ignoring the fact that with the exception of certain portions of Long 
Island there is almost no artesian capacity, and on Manhattan and the 
mainland the crystalline rocks make such development useless; (c) Lake 
Ontario, apparently overlooking the great distance (400 miles) and the 
many other complications that this international water body involves; 

(d) the Housatonic river, neglecting the difficulties of interstate origin ; 

(e) Dutchess county, where the city is prohibited by legislative enactment; 
(0 the Hudson river, ignoring the fact that the Hudson is an estuary 
of the sea with brackish water of a very impure quality and wholly unfit 
for domestic uses. It is, however, worth while to note that Hudson river 
water is sure to be used more and more extensively for fire protection and 
similar purposes in the more densely populated portions of the city by 
means of an entirely different system of conduits. This is one of the 
most promising directions of relief looking to the more distant future. 


I I 

The Catskill supply will furnish over 500,000,000 gallons of 
. water daily and was estimated to cost $161,857,000. That is, the 
additional supplies from the Catskills as planned will, when com- 
pleted, be sufficient for the increasing demands of the growing city, 
; for the next 35 years. And some of it may be badly needed long 
before it can possibly be delivered. 

Parts of the Catskill system 1 

The chief sources within the Catskills now included in the plans 
of the board are : 

1 Esopus creek, to be taken at a point near Olive Bridge. 

2 Rondout creek, to be taken at a point near Napanoch. 

3 Three small Streams tributary to the Rondout. 

4 Schoharie creek, to be taken at a point near Prattsville. 

5 Catskill creek, to be taken at a point about 1 mile northeast of 

6 Six small streams tributary to the aqueduct between Catskill 
creek and Ashokan reservoir. 

The comparative areas of watershed and their daily capacity are 
estimated 2 by the corps of engineers as follows : 




1 Esopus watershed 

2 Rondout watershed . . . 

3 Three small tributaries 

4 Schoharie watershed . . . 

5 Catskill watershed 

6 Six small streams 





70 000 000 000 s 

20 000 000 000 

45 000 000 000 

30 000 000 000 

250 000 000 

98 000 000 

27 000 000 

136 000 000 

100 000 000 

49 000 000 


165 000 000 000 

660 000 000 

1 The subdivisions and proposed locations given here are taken chiefly 
from the Report of the Board of Water Supply of the City of New York 
to the Board of Estimate and Apportionment, October 9, 1905. 

2 Estimates are much more complete for the Esopus, which it is planned 
to develop first, than for any other streams ; and it must be understood 
that the figures are subject to revision dependent upon modifications of 
original plans to meet the conditions that develop upon more elaborate 

3 Preparations are to be made for storage of 120,000,000.000 gallons of 
water on the Esopus, but a part of this capacity is intended to accommodate 
supplies drawn from other sources than Esopus creek itself. 



The evident certainty that present supplies from the Croton and 
Long Island will be very inadequate long before the Catskill system 
can be completed has influenced the adoption of plans contemplating 
the construction of certain parts in advance of the rest. To begin 
with, only the Esopus watershed is to be developed by the con- 
struction of the great Ashokan dam at Olive Bridge making the 
reservoir of full capacity. At the same time that portion of the 
aqueduct between the Ashokan dam and the present Croton reser- 
voir is to be completed in advance of other parts so as to make it 
possible to turn additional supplies into the Croton system, the 
capacity of the present Croton aqueducts being somewhat in excess 
of the Croton storage in dry years. It is furthermore desirable that 
increased storage capacity should be secured much nearer to New 
York city, and with that end in view Kensico reservoir is to be 
greatly enlarged. It is estimated that this may be made to hold 50 
days' supply of 500,000,000 gallons daily. 

The development of the Catskill system is being carried on by the 
Board of Water Supply, which was appointed by Mayor McClellan, 
as provided in chapter 724, of the laws of 1905. The present board 
consists of John A. Bensel, president, Charles N. Chadwick and 
Charles A. Shaw. The Engineering Bureau of the Board is in 
charge of J. Waldo Smith, as chief engineer, Merritt H. Smith, as 
deputy chief engineer and Thaddeus Merriman, assistant to chief 

Influenced doubtless in large part by the unity of certain portions 
of the project, either because their essential engineering features 
are distinct, or because their construction is more urgent, or in order 
to facilitate the work of supervision of so great an undertaking, the 
following departments have been created : 

1 Headquarters department (executive). In charge of general 
designs, plans of construction and preparation of contracts. Alfred 
D. Flinn, department engineer. 

2 Reservoir department. In charge of development of the Cats- 
kill watershed and the construction of the various dams and res- 
ervoirs. Carlton E. Davis, department engineer. 

3 Northern aqueduct department. In charge of the construction 
of full capacity aqueduct from the Ashokan dam (60 miles) to Hunt- 
ers brook in the Croton system. Robert Ridgway, department engi- 

4 Southern aqueduct department. In charge of the construction 
of full capacity aqueduct from Hunters brook in the Croton system 



to Hill View reservoir on the northern limits of New York city 
and of the storage reservoirs and filtration work. Merritt H. Smith, 
and more recently F. E. Winsor, department engineer. 

5 Long Island department, in charge of the development of the 
underground water supply of Long Island. A plan looking toward 
this end has been prepared and approved by the city authorities and 
is now being reviewed by the State Water Supply Commission. 

6 City aqueduct division. In charge of the delivery of water 
from Hill View reservoir throughout Greater New York. Origi- 
nally in charge of W. W. Brush, now under Walter E. Spear, as 
department engineer. 

Departments are further divided into " divisions " each in charge 
of a division engineer and a full corps of assistants. The subdi- 
visions of these larger units, although primarily based upon con- 
venience and efficiency of engineering supervision, coincides rather 
closely with the larger geologic problems included in this bulletin. 


Generalities of construction 

The chief types of structure projected include (i) masonry dams, 
(2) earth dikes with core walls, (3) "cut and cover" aqueduct 
through country of about the elevation of hydraulic grade, (4) 
tunnels through mountains or ridges that are too high, and (5) 
pressure tunnels under valleys or gorges that are too low. 

Some of these are of record proportions. For some of the de- 
tails and figures see the different special problems in part 2. 

All items complete as planned involve a total of: 

10 dams 

10 impounding, storage and distributing reservoirs 
4.5 miles of dikes 

54.5 miles of "cut and cover" aqueduct 
13.9 miles of tunnel at grade 
17.3 miles of pressure tunnel below grade 
34 shafts of aggregate depth of 14,723 feet. 
6.3 miles of steel pipes making 

92.5 miles of aqueduct complete to Hill View equalizing 
1 filtration works 

18.0 miles of delivery tunnel in New York city to the terminal 

shafts in Brooklyn 
16.3 miles of delivery pipe lines 



Allowing for contingencies and costs for engineering supervision 
the system is estimated to cost $176,000,000 and many years will 
be required for its completion. The present plans, however, con- 
template only the immediate development of the Esopus watershed, 
the storage reservoirs near the city and the main aqueduct to the 
various points of delivery within the city limits. It is expected 
that part of this additional supply of water will be available by the 
year 1913, or early in 1914. 



W'hen the Ashokan reservoir is filled the surface of the stored 
waters will stand 590 feet above the sea. Hill View reservoir on 
the northern borders of New York city will have an elevation of 
295 feet. The distance between these two points is nearly 75 miles 
in direct line. The contour of the country and other exigencies 
of construction will increase this to approximately 92 miles. A 
main distributary conduit in New York city will add 18 miles more. 

The destination of the water therefore before distribution begins 
is 300 feet lower than its starting point. This is sufficient head to 
permit gravitational flow and a self-delivering system. If the hy- 
draulic gradient can be maintained it would evidently constitute a 
decided advantage. The plans have therefore from the beginning 
contemplated such construction. It means then that a flowing 
grade must be maintained in all tunnels or channels or tubes and 
that when a depression has to be crossed the pressure must be 
maintained in some sort of a conduit so that the water may rise 
again to a suitable level on the other side. 

The difficulties of accomplishing this in a work of such magnitude 
are not at first apparent. The full significance of the undertaking 
can be realized only after a study of the country through which the 
aqueduct must be carried. It then resolves itself into a series of 
problems, each one having its own characteristics and peculiar 
difficulties and methods of solution and each requiring a thorough 
understanding of the topographic features of the vicinity and a 
working knowledge of geologic conditions. 

General questions 

It is sufficient at this point to call attention to the facts of the 
topographic map and point out only the most general physiographic 
features that may at once be seen to materially modify the simplicity 
of the line. 

For example, one has scarcely left the great reservoir, with water 
flowing at 580-90 feet above tide, before the broad Rondout 
valley is reached, with a width of 4*/. miles nowhere at great 
enough elevation to carry the aqueduct at grade. If it is to be 
crossed at all, and it must be crossed to reach New York city, some 




special means must be devised. If a trestle be proposed, one finds 
that it would have to be 4>j miles long (24,000 feet), and in some 
places 300 feet high, and at all points large enough and strong 
enough to carry a stream of water capable of delivering 500,000,000 
gallons daily — a stream that if confined in a tube of cylindrical 
form would have a diameter of about 15 feet. 

A steel tube might be laid to carry the water across and deliver 
it again at flowing grade, but here one is met with the fact that it 
would require a tube of unprecedented size and strength and if 
divided into a number of smaller ones the cost would be greater than 
that of a tunnel in solid rock. 

The other alternative is to make a tunnel deep enough in bed 
rock to lie beneath surface weaknesses and superficial gorges and 
in it carry the water under pressure to the opposite side of the 
valley. This is the plan that seems best suited to the magnitude 
of the undertaking and would seem to promise most permanent con- 
struction. But no sooner is this conclusion reached than it is 
realized that there are now several hitherto unregarded features 
that assume immediate and controlling importance. Some of these, 
for example, are (1) the possibility of old stream gorges that are 
buried beneath the soil, (2) the position of these old channels and 
their depth, (3) the kinds of rock in the valley, (4) their character 
for construction and permanence, (5) the possible interference of 
underground water circulation, (6) the possible excessive losses of 
water through porosity of strata, (7) the proper depth at which the 
tunnel should be placed, (8) the kinds of strata, and their respective 
amounts that will be cut at the chosen depth, (9) the position and 
character of the weak spots with an estimate of their influence on 
the practicability of the tunnel proposition. Then after these have 
all been considered the whole situation must be interpreted and 
translated into such practical engineering terms as whether or not 
the tunnel method is practicable, and at what point and at what 
depth it should cross the valley, and at what points still further 
exploration would add data of value in correcting estimates and 
governing construction and controlling contracts. 

This is a general view of one case, the first one of any large 
proportions in following down the aqueduct. • There are many 
others. In nearly all of them the importance of geologic questions 
is prominent. Many of them, of course, are of the simplest sort, 
but, on the other hand, some are among the most obscure and 
evasive problems of the science. And they do not become any 




easier simply to know that they must ultimately be stated in terms 
precise enough for the use of engineers, and to know furthermore 
that the real facts are to be laid bare when construction begins and 
as it progresses. But from another viewpoint it may be regarded 
as an exceptionally fine opportunity to study applied geology in its 
best form and to see the intimate interrelationship between an 
engineering enterprise of great public utility and a commonly con- 
sidered more or less obscure science. The services of geology have 
been seldom so consistently employed in earlier undertakings of 
similar character. It is to be hoped that the accompanying illus- 
trations of the practical application of geologic knowledge and facts 
to engineering plans and practice may add to the appreciation of 
the commonness and variety of such service in many everyday 
affairs. Furthermore, this unique enterprise, the like of which for 
magnitude and complexity has never before been attempted, has 
given to those whose good fortune has brought them into working 
relations with its problems, the opportunity of a generation in their 
chosen field. 1 The success stages from isolated observations, 
inference, hypothesis, theory, conclusions, and fully proven facts 
are all represented. The steps more or less fully coincide with the 
degree of confidence observable in the tone of advisory reports to 
the engineers in charge — representing suggestions, recommenda- 
tions, or specific advice. 

Tt is one of the cherished wishes of the writer of this bulletin 
that some of these problems may be presented in such manner as 
to serve a distinct educational purpose. For this reason in part, 
deeming it even of greater importance than the mere enumeration 
of newly discovered facts, the writer has chosen to treat the sub- 
ject from the standpoint of an instructor illustrating the develop- 
ment of working conclusions. Tt is certain that not all readers have 
the same degree of preparation or acquaintance with the subject- 
matter, and it may therefore be useful to include many things that 
some may well pass by. No excuse is offered except that such 
method of treatment, in behalf of the general intelligent public that 
it is hoped to reach, seems to the author to be advisable. 

1 W. O. Crosby of the Massachusetts Institute of Technology, James 
F. Kemp and Charles P. Berkey of Columbia University have constituted 
tlie staff of consulting- geologists throughout most of the exploratory work. 



Other problems 

The foregoing observations apply likewise to the other larger 
problems of the aqueduct line. A list of the larger ones requiring 
extensive exploration and illustrating geologic application in their 
solution are given below : 

1 Location of the Ashokan dam 

2 Sources of material for construction 

3 Crossing the Rondout valley 

4 The Wallkill valley 

5 Moodna buried valley 

6 Pagenstechers gorge and Storm King mountain 

7 The Hudson river crossing problem 

8 The Storm King-Break Neck cross section 

9 Foundry brook 

10 Sprout brook notch 

1 1 Peekskill creek valley 

12 Croton lake pressure tunnel 

13 Bryn Mawr siphon 

14 The new Kensico dam 

15 Kensico quarries 

16 New York city delivery tunnel 

In addition to these there are several questions of general bear- 
ing in which the chief lines of argument and the chief basis of con- 
clusion are essentially geologic. Although little wholly new data is 
yet available on these particular questions from any direct work of 
the aqueduct, yet it will add materially to an appreciation of the 
far-reaching influence of established geologic data and geologic rea- 
soning to enumerate some of them : 

17 Continental subsidence and elevation 

18 Crustal warping 

19 Postglacial and present faulting 

20 Underground water circulation 

21 Relative resistance of the different formations to corrosion by 
aqueduct waters 

22 Structural materials 

Each of these problems or questions or topics is discussed sepa- 
rately, so far as practicable. By adopting this plan, of course there 
is a tendency to repetition but this to a certain extent is unavoid- 
able. Some of it is overcome by suitable references to preceding 



discussions. Where such cross reference is too cumbersome, the 
items are repeated in preference to leaving the case obscure. Thus 
it is hoped to make each case a unit, and the whole series useful 
and understandable. 

Gathering data 

In the accumulation of data all the members of the engineering 
corps 1 as well as the men acting only in a consulting capacity have 
taken part. Necessarily the bulk of the exact data has been 
gathered by the men all the time on the ground and whose duty it 
was to superintend explorations. The care and intelligence with 
which this has been done is notable. A considerable proportion of 
the labor of manipulating the accumulated data and interpreting it 
so as to reach an explanation of conditions and formulate conclu- 
sions has been assumed by the consulting men. 

Too much credit can not be given to the heads of departments 
and divisions for the open-handed way in which all needed facts 
were held available at all times for comparison and guidance toward 
sound conclusions. The information upon which investigations 
have been initiated have been chiefly the following: 

1 The geologic maps and reports of the New York State Survey 

2 United States topographic maps 

3, Geologic folio no. 83, New York city folio 

4 Earlier engineering records and reports 

5 Reports of special commissions on water supply 

1 In this work, no group of men have had so direct responsibility as the 
division engineers. The success with which so many complicated explora- 
tions were carried out is chiefly due to their constant care and foresight 
and perseverance and the able assistance of their staff. Those who have had 
especially important divisions for the geological problems involved are given 
due credit in the discussions of part 2, of this bulletin. It is easy, how- 
ever, to neglect sufficiently full acknowledgment of their services in gather- 
ing and formulating data of this kind. Among those having charge of the 
most important exploratory work the following names should appear: 

James F. Sanborn, for sometime assigned to geologic work on the North- 
ern aqueduct. 

William E. Swift, in charge of the Hudson river explorations. 
William W. Brush, in charge of the early New York city explorations. 
Lazarus White, in charge of the Rondout valley explorations. 
Lawrence C. Brink, in charge of the Wallkill division explorations. 
J. S. Langthorn, in charge of the exploratory work at the Ashokan Reser- 

Wilson Fitch Smith, in charge of work at Kensico dam, and 
A. A. Sproul, in charge of the Peekskill creek and Sprout brook explora- 



Some of these are printed reports and records not directly con- 
cerned with this enterprise, but whose' information has been found 
useful in this field. This is especially true of the first four sources 
enumerated, i, 2, 3, 4. The last is a specific study with direct 
reference to this project. 

Investigations were begun from the above vantage point. The 
methods employed and the explorations conducted constituting the 
further sources of information and furnishing the complete data 
upon which all conclusions have been based include the following: 

6 Detailed topographic studies of the engineers of the Board of 

Water Supply 

7 Geologic field work making observations in detail of all geo- 

logic factors that seem to bear on the problem in hand 

8 Wash borings for depth to bed rock 

9 Chop drill holes through stony ground to bed rock 

10 Shot drill holes in bed rock 

11 Diamond drill holes 

12 Test pits and trenches for detail of drift structure 

13 Test tunnels in rock for working quality 

14 Deflection tests for holes that have swerved aside 

15 Pumping tests for underground water supply 

16 Pressure tests for rock porosity 

17 Microscopic examinations of rock types 

18 Laboratory tests of quality and behavior of materials. 

The mass of data accumulated from all these sources is surpris- 
ing. For example, there are upward of 200 wash borings on the 
different proposed Hudson river crossing lines alone ; there are 69 
drill borings and 177 wash borings on the site of Kensico dam; 
there are 69 shot and diamond drill holes on the Rondout siphon 
line aggregating 10.234 feet of rock core ; there are 65 drill holes of 
various sorts on the Moodna creek siphon aggregating in total pene- 
tration of drift over 10,000 feet: there are 106 borings, besides 
several pits and trenches at Ashokan dam location. At every point 
explorations suitable to the particular problems in hand were con- 
ducted. The whole mass of data is conveniently recorded, much of 
it is tabulated, some of it is represented graphically, samples of 
nearly all of the material are available for examination. 1 and all 

1 The cores of all drillings and suitable samples of all borings in drift 
have been saved and properly labeled and are to be permanently boused at 
some convenient point on tbe aqueduct line when completed. At present 
they are cared for at the different division offices. 



have been made use of in coining to a consistent understanding of 
the conditions. 

But the amount of accumulated data is no more remarkable than 
the difficulties that have been encountered in obtaining it. For 
example, in the Moodna valley it has taken three to four months' 
time to put down a single hole to bed rock — the average time con- 
sumed for each of the 15 holes exploring the deepest portion of 
the valley was about 60 days. The chief trouble is caused by heavy 
bouldery till. In one case a boulder was penetrated for 35 feet, 
lying a hundred feet above bed rock. 

The extreme of such difficulty is, of course, encountered in the 
Hudson river itself, where the drill has to contend with: (1) the 
rise and fall of the tides, (2) the river currents, (3) a maximum of 
90 feet of water, approximately 700 feet of silt, gravel, till, boulders, 
etc., filling the old preglacial gorge. The heavy steamboat and 
towing traffic has been a serious element in the problem. Probablv 
never anywhere have drillmen had to face so nearly insurmount- 
able obstacles. In two years only two holes reached below a depth 
of 600 feet below sea level. A third, now in progress, has pene- 
trated a depth of 768 feet without entering rock. 



In the earlier stages of work topographic features were of most 
concern, and they largely controlled the selection of reservoir sites 
and possible lines for the aqueduct to follow. It was, however, 
at once recognized that tunnels would be unavoidable and studies 
as to the types of rock formations to be encountered were begun 
It was also early appreciated that the soil or drift cover is very 
unevenly distributed over the rock surface and that, especially in 
the chief valleys requiring pressure tunnels, it would be necessary 
to determine the profile of the rock floor. At this point wash bor- 
ings were begun. But the natural limitations of the wash rig 1 for 
penetrating drift of all kinds left the information still too indefinite. 
The wash rig can not penetrate hard rock. It can not wash up 
anything but the finer matter, and a boulder of very moderate size 
is almost as effectual a barrier as true rock ledge. By a combination 
of washing and chopping or by the use of an explosive to break or 
dislodge an obstruction some progress in unfavorable material may 
be made, but the wash rig alone, in a drift-covered region, gives 
only negative results. It is certain, for example, that bed rock lies 
at least as deep as the wash rig has penetrated, but it is not certain 
that it is bed rock instead of some other obstruction. Except in 
areas of special drift conditions, 2 therefore, the wash rig was insuf- 
ficient. To rely upon the process at random was clearly impossible, 
and to determine whether or not the results of a particular locality 

1 A " wash rig " is a device composed essentially of two iron pipes, one 
within the other, and so mounted that the inner one can be worked up 
and down in sort of a churning fashion while water under considerable 
pressure is forced through it to the bottom and out again by the larger 
pipe to the surface, carrying up with the current the displaced sand and 
clay. As progress is made with the inner pipe the outer one is from time 
to time driven down and the process renewed and repeated till the hole is 

2 One of the most notable areas of special drift conditions is repre- 
sented in the Walkhill valley Isee discussion in pt 2] where there were 
developed large deposits of modified drift, stratified gravel, sand and clays, 
lying immediately upon the bed rock floor. In this the wash bore 
process was eminently satisfactory, and the rapid progress made by it 
together with its economy made this an especially attractive method of 




cuiild be relied upon became involved at once with an interpretation 
of local glacial phenomena, especially an interpretation of the char- 
acter of the local drift. In order to see the limited application of 
this method one needs only to point out that the majority of drift 
deposits in this region are stony or even bouldery, forming thick 
coverings in the valleys, and to call attention to the experience at 
two or three points. For example, at Moodna creek, the prelimi- 
nary wash borings were obstructed and bed rock reported at 5 to 
15 feet below the surface where afterward, by other means, it was 
proven to lie more than 300 feet down. Or again, in the pre- 
liminary wash borings in the Hudson, the rigs were stopped and 
rock bottom provisionally reported at from 25 to 200 feet below 
sea level, but later explorations have proven at the same point that 
rock bottom is more than 700 feet down. 

Therefore, to the " wash rig " was added the " chop drill " and 
the " oil-well rig " and to these, or to modifications of them, 1 the 
success in reaching bed rock has been due. 

From independent field studies of a more strictly geologic nature 
it became clear that many of the valleys, where pressure tunnels 
were proposed, are of comparatively complex geologic structure and 
exhibit considerable variety of rock quality and condition. This 
then introduced and necessitated still more elaborate lines of ex- 
ploration. It was not enough to know the profile of rock floor 
alone, it became of equal importance to penetrate the rock and obtain 
samples of it. So the shot drill 2 and the diamond drill 3 were 
employed and the drill cores preserved for identification and 

1 The essential features of the machines in most instances are, a high 
tower or support, a heavy chisel-shaped plunger that can be raised by 
a rope and dropped repeatedly in the hole, destroying or displacing 
obstructions, and which can be followed by a casing driven down as 
progress is made — a combination of washing, chopping and driving. 

2 The shot drill, or calyx drill, is essentially a machine devised to rotate 
a steel tube which is so adjusted and manipulated that a supply of small 
chilled shot can be kept continually under the lower end as it bores into 
the rock. The cutting is done by the shot immediately under the edge 
of the tube. A core remains in the~ tube and may be recovered. Its best 
position is vertical. 

3 The diamond drill consists essentially of a bit or crown set with black 
diamonds (bort) in such manner that when the bit is attached to a rotating 
tube a circular groove is cut into the rock. By proper attachment to 
jointed tubes and driving gear a hole may thus be bored at any angle and 
to great depth and a core recovered. 



These preserved cores, now aggregating many thousands of feet 
have been of great service in determining the precise limits of 
formations and consequently the geologic structure or cross section, 
by which detailed estimates may be guided. 

Even these occasionally appeared to give insufficient data. The 
peculiar behavior of certain holes, as, for example, one or more 
at Foundry brook, 1 led to the suspicion that the drill had swerved 
from its course, following a particularly soft seam or zone, and 
that the results secured by it without large corrections, were wholly 
misleading. Tests proved that there had been a deflection. 

At this and many other places it later became very desirable to 
form some quantitative as well as qualitative opinion of the condi- 
tions existing in the underlying strata. The percentage of core 
saved, the rate of progress of the drill, the behavior of the drill, the 
condition of the core recovered, the loss of water in the hole — all 
these of course were considered. 

For more definite evidence as to porosity and perviousness, a 
series of carefully planned pressure tests 2 were made. By shutting 
off connection with the walls of the hole above a certain stratum 
and forcing water in under pressure, it was possible to demonstrate 
that certain strata or certain portions were practically impervious 
in their natural bed, while others were much less so, and to get an 
idea of their relative efficiency as water carriers. For the pressure 
tunnels, especially, this test is a very suggestive line of investigation. 

1 At Foundry brook \sce discussion of this problem in pt 2], the remark- 
able condition apparently shown was a reasonably substantial ledge of 
granitic gneiss, 50 feet, followed below by 200 feet of apparently soft 
sand and reported as such. No core could be recovered. So extensive 
a zone or bed or layer or mass is hardly conceivable considering the 
crystalline silicious character of the rock. Tt probably represents a steeply 
dipping crush zone along fault movement where the increased underground 
circulation has been unusually effective in producing decay. After enter- 
ins this zone the drill swerved from its initial course and kept within the 
soft seam. 

2 The pressure test is made by means of a force pump, fitted with a 
gage on which the pressure is recorded, connected by a pipe to the por- 
tion of the hole to be tested, and so adjusted to a device for blockading; 
or damming the hole that the water pressure is confined to those portions 
of the walls of the hole below the dam, or between two dams if an upper 
and lower one are used. In this way any portion of a hole, or stratum or 
several beds together may be tested and the amount of water absorbed 
per unit of time per unit of pressure determined. This is, of course, 
directly related to the porosity of the rock and is approximately inversely 
proportional to its presumed value as an aqueduct carrier. 



Where the strata are especially porous and where underground or 
permanent ground water supplies are very extensive and where at 
the same time the largest or deepest pressure tunnels are projected 
some uneasiness has been entertained as to the extent of interference 
from inflowing water during construction. An attempt to form 
some idea of the ease of such underground circulation has been 
made by a systematic pumping of one or two critical holes. The 
results leave many factors still too obscure to draw definite con- 
clusions. The test will be taken up again in the discussion of the 
Rondout siphon in part 2. 

Laboratory tests and experiments on materials complete the list 
of lines of investigation with which this bulletin is concerned. 
Although from the nature of the case these are elaborate and 
unusually complete, the more important lines are not at all new. 
All the methods of petrographic, chemical, and physical manipula- 
tion that seem to promise practical results of value to the success 
of the undertaking are followed and the data are organized and 
interpreted and conclusions are formulated with as great definite- 
ness for practical bearing as other lines of investigation. 



It will save much repetition and it is believed will altogether 
serve a useful purpose in maintaining unity of treatment to give 
an outline of the geologic features of the region in advance of the 
discussion of special problems. It is intended only for those not 
sufficiently familiar with the general geology to follow subsequent 

The region includes some of the most complicated and obscure 
sections of New York geology. It is simple in almost no one of the 
larger branches of the subject. In physiography there is the long 
and involved history and the results of long continued erosion of a 
variable series of formations in different stages of modification as 
to structure and metamorphism and attitude, modified still fur- 
ther by subsidences and elevations, depositions and denudations, 
peneplanations and rejuvenations, glaciation and recent erosion — 
all together introducing as much complexity as can well be found in 
a single area. 

In stratigraphy the whole range of the eastern New York geologic 
column is represented from the oldest known formation up to and 
including the Middle Devonic — a succession of at least 25 distinct 
formations which may for convenience be treated in groups that 
have had similar history. Each of these formations has a constant 
enough character to map and regard as a physical unit. Even this 
classification ignores the great range of petrographic variability 
shown in such formations as the Highlands or Fordham gneisses. 
All but two or three of these formations will be cut by the tunnels 
of the aqueduct. 

In petrography the range is even greater — so great, in fact, that 
only an enumeration of the variations will be attempted. They 
include elastics, metamorphics and igneous types ; stratified and un- 
assorted, coarse and fine, detrital and organic, marine and fresh 
water, homogeneous and heterogeneous, argillaceous, calcareous and 
silicious sediments, unmodified and thoroughly recrystallized strata ; 
acid and basic and intermediate intrusions ; massive and foliated 
crystallines — of many varieties or variations in each group. 

In tectonic geology an equal complexity prevails. There are regu- 
lar stratifications, cross-beddings, disconformities, overlaps and un- 
conformities ; interbeddings, lenses and wedges ; flat, warped, tilted 



and crumpled strata; monoclinal and isoclinal, open and closed, 
anticlinal and synclinal, symmetrical and overturned, horizontal and 
pitching folds ; joints, crevices, caves, crush zones, shear zones, and 
contacts; normal, thrust, dip, strike, large and small faults; veins, 
segregations, inclusions, dikes, sills, bosses and bysmaliths. 

With such variety of natural conditions it is not surprising that 
the problems of the aqueduct are also of great variety. No two 
have in all respects the same factors in control and no two can be 
explored and interpreted upon exactly the same lines. 

i Geographic features or districts. (Physical geography 1 ) 

It will be convenient at this point to think of the surface topog- 
raphy by districts — not wholly distinct from each other, but still 
with essential differences of origin and form. From south to north 
they are: (a) New York-Westchester county district. The area 
of crystalline sediments. South of the Highlands, (b) Highlands 
of the Hudson (Putnam county), (c) Wallkill-Newburgh district. 
From the Highlands to the Shawangunk range, (d) Shawangunk 
range and Rondout valley, (e) Southern Catskills. 

All have been sculptured by the same forces and with similar 
vicissitudes, but the difference of history and structure and condi- 
tion, already established when the physiographic forces began on the 
work now seen, have caused the variety of surface features indi- 
cated in the divisions made above. The more noticeable character- 
istics of these five districts are here given. 

a New York- Westchester district. The area south of the 
Highlands proper is characterized by a comparatively regular suc- 
cession of nearly parallel ridges separated by valleys of nearly equal 
extent (y 2 to 5 miles wide), making a surface of gently fluted 
aspect and of moderate relief (0-500 feet) sloping endwise toward 
the Hudson and the sea. The controlling factors in producing this 
topography are involved in a series of folded, foliated, crystalline 
sediments, of differing resistance to destructive agencies. 

b The Highland region is one of rugged features, with a 
range of elevation of 0-1600 feet A. T., forming mountain masses 
and ridges separated by very narrow valleys all having a general 
northeast and southwest trend across which the Hudson cuts its 
way in a narrow, angular gorge, forming the most constricted and 
crooked portion of its lower course. The bed rock is all crystalline, 

1 The physiographic history of a region is not understandable without a 
comprehensive knowledge of its geologic features and structures and history. 
Tt is therefore treated in a later paragraph. 



of massive and foliated types,, metamorphosed sediments in part 
with large masses of igneous intrusions and bosses. 

c The Wallkill-Newburgh district lying immediately north 
of the Highlands and extending to the Shawangunk range is a 
region of gently rolling contour. Most of the area along the pro- 
posed lines lies between 200 and 500 feet above the sea. There are 
only occasional rugged hills or short ridges, such as Snake hill and 
Skunnemunk. The valleys are broad and smooth and the divides 
are simply broad, hilly uplands. Bed rock is chiefly Hudson River 
slates with occasional belts of Wappinger limestone. The larger 
features, the trend of divides and valleys, are northeast and south- 
west, although this regularity is not so marked as in the preceding 
two districts. But the chief streams flow either northeast or south- 
west to the Hudson along these general lines. 

d The Shawangunk range and Rondout valley form a 
transitional unit from the complicated structural and tectonic con- 
ditions of the southerly districts to the uniform and almost undis- 
turbed strata of the Catskills. Its southeasterly half is a mountain 
ridge partaking of extensive faulting and folding and represented 
by the Hudson River slates overlain unconformably by the thick 
and very resistant Shawangunk conglomerate forming high east- 
ward-facing cliffs. Toward the northwest these disturbances dimin- 
ish, the strata gradually pass deeper beneath a great succession of 
shales, limestones, and sandstones of the Helderbergian series, and 
a broad valley is eroded in the softer portions. It is limited on the 
northwest by the prominent and very persistent escarpment border- 
ing the Hamilton series and forming the outer margin of the Cats- 
kill mountains. 

e The Catskill area is of simple structure. The strata are 
well bedded and lie almost flat with a gentle dip northwest. The 
surface features form a series of irregularly distributed escarp- 
ments, hills, valleys, cliffs, gorges and mountains, rising rapid!/ 
toward the west, with moderate to strong relief and reaching ele- 
vations of 2500 feet. The failure of the northeast-southwest trend 
of feature that is so common in all of the other districts is a marked 
difference. It is directly due to the flatness of the strata. 

2 Stratigraphy 

There are no strata of prominence in association with the main 
aqueduct younger than Devonic age except the glacial drift. Imme- 
diately adjacent areas, however, some of which are covered by the 
accompanying maps, and Long Island have later formations ex- 



tensively developed. Such are the Triassic rocks of the New Jersey 
side of the Hudson below the Highlands, and the Cretaceous and 
Tertiary strata of the Atlantic margin on Long Island and Staten 
Island. The development of underground water supply on Long 
Island is especially concerned with these later formations, and with 
the modified drift deposits of the continental margin. 

The whole series of formations are more commonly considered 
in groups that exhibit certain age or physical unity and that are 
for the most part characteristic of certain regional belts and that 
coincide somewhat roughly with the physiographic divisions already 
noted. There is in the following description and tabulation no direct 
attempt to unduly emphasize this relation or to belittle the divisions 
recognized in the commonly adopted geologic column. It is, how- 
ever, for the purpose in hand, more convenient and useful to keep 
clear the physical groupings, because largely these groups, instead 
of the more arbitrary subdivisions of age, are the units used in con- 
sidering structural and applied problems. 

a Quaternary deposits, (i) Glacial drift. A loose mantle of 
soil and mixed rock matter covers the bed rock throughout the 
whole region except (a) here and there where the rock sticks up 
through (outcrops), and (b) at the most southerly margin along 
the coast where the glaciers seem not to have reached. 

Origin. This mantle is usually very different in lithologic charac- 
ter from the underlying rock floor. There is almost always an 
abrupt break between the rock floor and the overlying material. 
The rock floor is grooved, smoothed, and scratched as if by the 
moving of rock or gravel over it. The larger boulders are usually 
of types of rock identical with ledges lying northward at greater or 
less distance. Materials of exceedingly great, variety both in 
size and condition and lithologic character are often all piled to- 
gether in the most hopelessly heterogeneous manner. These are 
now commonly regarded as conclusive evidence of glacial origin. 
There is no need of making the discussion exhaustive. It is almost 
universally called the " drift." 

Thickness. The thickness of the drift varies from almost o to ap- 
proximately 500 feet. It is generally thickest in the valleys where it 
has simply filled many of the original depressions and obliterated 
much of the ruggedness of surface, the gorges and ravines and can- 
yons of the preglacial time. 

Sources. It appears from an examination of the grooves and 
striae on bed rock, and the relationship of the different types of 
drift to each other, and from a comparison of the types of boulders 



with the ledges that may be regarded as their source, that the gen- 
eral ice movement was from north to south swerving along the 
southerly extension to east of south. Therefore it is not unusual 
to find abundant boulders of Palisade trap stranded in New York 
city or on Long Island, or boulders of the Cortlandt series, or of 
the gneisses of the Highlands, or, in occasional instances, of sand 
stones from the Catskills, or the limestones from the Helderbergs 
or perhaps in rarer cases even rocks from greater distance, as the 
Adirondack mountains. 

Kinds of drift. There are in the region two fundamentally differ- 
ent types of drift as to method of deposition. They are (a) unas- 
sorted drift (till or hardpan), and (b) modified drift (stratified or 
partially assorted gravels, sands, clay, etc.). The former (a) repre- 
sents deposition directly from the ice sheet at its margin (terminal 
or marginal moraines) or beneath ("ground moraine") without 
enough water action to rework and assort the material. It there- 
fore contains boulders, pebbles, sand and clay of a heterogeneous 
mixture of the most complex sort both as to size and character. In 
such deposits there is almost always sufficient intermixture of clay 
and rock flour of the finest sort to make a very compact and dense 
mass that is usually quite impervious to water. Such deposits are 
distributed rather unevenly over the surface and where this uneven- 
ness leaves hollows or basins, or obstructs the outlets of other de- 
pressions, they may hold water and form small lakes or ponds or 
swamps. This is almost universally the origin of the many thou- 
sands of lakes of the northern lake region. It is evident that ma- 
terial of this character, a type that commonly serves the purpose of 
a natural dam or reservoir, would be especially important and useful 
at certain places on the Catskill system. As a matter of fact, so far 
as geologic features are concerned, it is the chief factor in choice of 
location for the Ashokan dam [see discussion pt 2] and is a con- 
trolling factor in the plans for the erection of the miles of dikes 
at less critical margins of the reservoirs. Till is an extensively 
developed type but frequently passes abruptly either laterally or 
vertically into assorted materials of very different physical char- 

(b) All materials associated in origin with the glacial occupation 
that have been materially modified especially in the direction of an 
assorting of material are referred to as " modified drift " deposits. 
They include (1) deposits made by both water and ice together, 
(2) those formed by running water. (3) those laid down in stand- 



ing water. Or again (i ) those accumulated rapidly with very irreg- 
ular supply of material at the margin of the ice-forming, hummocky 
or hill and kettle surface (kames, eskers), (2; those carried along 
valleys or general lines of drainage to a considerable distance beyond 
the ice margin aggrading the valley with the overload of gravels 
and sands (valley trains), (3) those washed out from the ice margin 
in more even distribution forming a gently sloping and thinning 
extramarginal fringe (outwash or apron plains), (4) those fine 
matters that are carried by glacial streams into the margins of more 
quiet waters, either a temporary or a permanent lake or a larger 
and slower stream or other body forming more perfectly assorted 
and more evenly stratified deposits (delta deposits), (5) those 
finer rock flours and clays that remain suspended longer and carry 
out much farther settling only in the very quiet waters of lakes 01 
estuaries or temporary water bodies of this character forming the 
perfectly banded clays (glacial lacustrine clays). 

It is evident then that modified drift has in the process of its 
accumulation suffered chiefly a separation of fine from the coarse 
particles and that in most cases the fine clay filling that makes the 
till dense and impervious to water, has been washed out and de- 
posited by itself in the more inaccessible deeper waters. As a re- 
sult most modified drift deposits are pervious and easy water 
conductors, but poor or questionable ground for dikes or dams or 
basins [sec discussion of Ashokan dam, pt 2]. 

Some of them, the medium sands and gravels, furnish an excel- 
lent and already cleaned structural material for concrete or mortar, 
such as the Horton sand deposit, or coarser kinds may be crushed 
and sued before using as is done at Jones Point on the Hudson. 

The finer silts and clays, usually overlain by assorted sands, are 
abundant along the Hudson, having been deposited there at a time 
when the water of this estuary stood 50 to 150 feet higher than 
now. Recent erosive activity of the river has cut the greater pro- 
portion of the original deposits away but at many places large quan- 
tities still remain above water level in the banks and still greater 
quantities extend beneath the river. These deposits are the support 
of the brick industry of southeastern New York. The till deposits 
are very difficult to penetrate in making borings because of the 
boulders, the wash rig being almost useless. Modified drift of the 
medium and finer sorts is easily and cheaply penetrated, and, if it 
lies on bed rock, such exploration gives reliable results. 

Structure. But this is stating the actual conditions too simply. 
The glacial epoch was a complex one The continental ice sheet may 



have advanced and retreated repeatedly, how many times in this 
region is not clear. \\ ith each time of advance and retreat, the 
work done by it partly destroyed, or disturbed or modified or cov- 
ered the earlier ones in what appears now to be a most arbitrary 
way (in reality, of course, in a very consistent way for the condi- 
tions that then existed). So one frequently finds a till beneath a de- 
posit of stratified drift, or modified drift beneath till, or a succession 
of a still greater number of changes in almost hopeless confusion. 
In New York city, for example, at Manhattanville cross valley, 
the exposed drift above street level includes (a) at the bottom, 
water-marked stony till and assorted gravels, (b ) in the middle per- 
fectly horizontal, stratified rock flour and the finest sand, (c) top, 
wholly unassorted bouldery till, covered by thin soil. It is evident 
that the most careful and accurate identification of the surface type 
without subsurface investigation would give, for such uses as are 
now being considered, thoroughly unreliable evidence as to the 
behavior of the whole body at this point. Therefore, a determina- 
tion of the changes and quality forms an essential record. All of 
these types are to be found in the region, but the different grades of 
till and roughly modified material belonging to the kame type are 
more common inland. 

On Long Island the development of marginal modified types is 
extensive and more or less obscured by the advance and retreat 
noted above. The larger divisions recognized in deposits are (a) 
an early accumulation of sands and gravels, strongly developed near 
the western end of the island, known as the " Jameco " gravel, 
(b) an interglacial (retreatal) deposit of blue clays known as the 
" Sankaty " beds, (c) a later series of deposits, sands, clays, gravels 
and till, belonging to the closing stages of the ice period correspond- 
ing to the surface deposits of the larger portion of the whole region 
(Tisbury and Wisconsin advances). Some of these sands and 
gravels are important water-bearing sources for the new Brooklyn 
additional supply. 

The whole Long Island series according to Veatch 1 includes : 

Tisbury stage <J 

Wisconsin stage 

1 After PP 44, U. S. Geological Survey, p. 33. 



Gay Head I Foldin S (g lacial folding) 

\ Sankaty retreatal stage (interglacial) clay beds 

Jameco <f Glacial - Jameco gravels 

L rostmannetto erosion (interglacial) 

Mannetto stage Glacial — old gravels 

A radically different and in some respects a much simpler inter- 
pretation 1 of the Long Island deposits has been outlined by W. O. 
Crosby. The essential feature of his classification is the unity and 
simplicity of the glacial epoch. Only the moraines and associated 
sands and gravels of outwash origin during advance and retreat are 
regarded as glacial. All other deposits below and including the 
Sankaty clay beds he regards as preglacial. 

The Jameco gravels are interpreted as Miocene in age. 

Certain persistent yellow gravels overlying the Jameco are classi- 
fied as Pliocene. 

b Tertiary and Cretaceous deposits. (2) Tertiary outliers. 
Deposits of Pliocene age are littoral in type [PP 44 U. S. G. S. 
p. 28] and are not very well differentiated (Long Island, Staten 
Island). Probably equivalent to the Bridget on beds of New Jersey. 

Certain " fluffy " sands in thin beds are assigned by Mr Veatch 
to the Miocene (Long Island, Staten Island). Probably equivalent 
to the Beacon hill deposits of New Jersey. Crosby places the 
Jameco gravels in the Miocene together with the Kirkwood lignitic 
and pyritic clays and sands. 

(3) Upper Cretaceous deposity 2 are extensively developed. 
They form the chief bed rock of Long Island. 

1 The writer offers both of these outlineG of the glacial and associated 
deposits in preference to either alone. Both Veatch and Crosby have given 
immensely more time to the study of these questions than any one else. 
It is hardly fitting for a newcomer in their field to reject either view. But 
because of the very great difference between the two interpretations one 
may be pardoned a preference. It is the writer's opinion that the simpler 
outline is the more tenable. It does not seem possible to establish a very 
complex series of stages in the glacial epoch as represented in the deposits 
of southeastern New York. 

2 Crosby's classification of the Cretaceous is as follows : 

(a) Monmouth — slight development of marls. (Lower and middle 

marl series.) 

(b) Matawan — (clay marl series) probably present on Long Island. 

(c) Magothy — an extensive series of variegated and micaceous sands 

and clays. Heavy development on Long Island. 

(d) Raritan — Plastic clay scales and the Lloyd sand. 



(o) A lignitiferous sand with occasional clay beds forming the 
uppermost of the Cretaceous series is probably equivalent to the 
marl series of New Jersey. But it lacks the prominent greens and 
development characteristic of the region further south. Not clearly 
separable from the underlying formation or Matawan beds. 

(£?) The Matawan beds. Gray sands and clays. 

(c) Raritan formation. Clays and sands, plastic clays, the Lloyd 
sand, an important water carrier lies about 200 feet below the top 
of the formation. Occasional leaf impressions. 

All of these formations, except where disturbed locally by glacial 
ice, dip gently seaward. The sand beds of these strata are the chief 
sources of underground water being developed by the new system. 

c Jura-Trias formations. (4) Palisade diabase. This is a thick 
intrusive sheet, or sill, of igneous rock of diabasic type. It is 
700-1000 feet thick. It lies for the most part parallel to the bed- 
ding of the surrounding, inclosing, sedimentary rocks, and, rising 
gently eastward, forms a strong cliff continuously along the west 
bank of the Hudson for 40 miles. It varies from very fine to very 
coarse texture and is for the most part fresh, tough, durable, and 
is the source of large quantities of the most satisfactory quality of 
crushed stone now on the market for use in concrete. 

(5) Newark scries. This is a very great thickness of silicious 
sediments, chiefly reddish conglomerates, red and brown quartzose 
and feldspathic sandstones and shales. They dip gently westward 
and northwestward at 10-20 degrees, and are confined, in this 
region, to the west side of the Hudson south of the Highlands. The 
formation supplies " brownstone " for building purposes. 

None of the Jura-Trias rocks, so far as known, will be cut by the 

d Devonic strata. (6) Cat skill formation. This formation 1 is of 
continental type, chiefly a conglomerate. A white conglomeratic 
sandstone forming the uppermost portion attains its greatest thick- 
ness on Slide mountain (350 feet). It is a " coarse grained, heavy 
bedded, moderately hard sandstone containing disseminated pebbles 
of quartz or light colored quartzite, and streaks of conglomerate." 

A red conglomeratic sandstone constitutes the much thicker por- 
tion below (1375 feet). It is a " coarse, heavy bedded sandstone of 
dull brownish hue containing disseminated pebbles and conglom- 
eratic streaks, differing from the overlying beds chiefly in color. In 

1 Grabau, A. W. N. Y. State Mus. Bui. 92. Geology and Paleontology 
of the Schoharie Valley. 



both series the pebbles and conglomeratic streaks are scattered and 
irregular, while the sands are often cross-bedded. Thin layers of 
red shale occur, and locally gray sandstones." The deposits prob- 
ably represent Hood plains, deltas, and alluvial fans accumulated 
mostly above sea level. 

(7) Onconta sandstone (Upper flagstone). "Thin and thick 
bedded sandstones from 20 to 200 feet thick with interbedded red 
shales up to 30 feet thick." Chiefly light gray to brown in color. 
Abundant cross-bedding, occasional dark shale, frequent flagstone 
beds. Capable of furnishing " bluestone " flags and more massive 
dimension stone. To be seen in the vicinity of West Shokan and 

(8) Ithaca and Sherburne (lower flagstone " bluestone "). " Thin 
bedded sandstone, with intercalated beds of dark shale. The sand- 
stones are in masses from a few inches to 40 feet in thickness, 
greenish gray to light bluish gray or dark gray in color, and are 
extensively quarried as flagstones." There are occasional conglom- 
eratic streaks. Occurs in large development in the vicinity of the 
Ashokan reservoir (500 feet). The heavier cross-bedded and 
coarser grained beds are capable of furnishing an unusually good 
quality of large dimension stone for heavy structural uses. The 
beds of this formation near Olive Bridge will in all probability 
furnish the greater proportion of stone of all kinds for the con- 
struction of the great Ashokan dam [see discussion of bluestone 
near Ashokan dam, pt 2]. The chief common fossil content is 
impressions of plant remains. 

(9) Hamilton and Marccllus shales. " Dark gray to black or 
brown shales with thin arenaceous beds in the upper part." Forms 
the upper portion of the escarpment that follows the outer margin 
of the Catskill foothills bordering the westerly side of the middle 
Rondout and lower Esopus valleys. Occasionally beds are sub- 
stantial enough for flagstone production (700 feet or more with the 

The chief index fossils are: Spirifer mucronatus, 
Athyris spiriferoides, Chonetes coronatus. 

The Marcellus shale is not readily differentiated in the Esopus 
valley Characteristically it is a thin bedded shale of no great 
thickness (180 feet in the Schoharie valley) lying between the 
Onondaga limestone and the Hamilton and obscured by talus from 
the escarpment (with the Hamilton 700 feet.) 

Styliolina fissurella, Chonetes mucronatus, 
Strophalosia truncata, Liorhynchus mysia. 



The dividing lines between the different sandstones and shale 
formations, the Oneonta, Ithaca, Sherburne, Hamilton and Mar- 
cellus, can not be sharply drawn in the Ksopus region. Together 
they form in a large way a rather satisfactory held unit. For 
specific purposes it is necessary to recognize that the lower por- 
tions are prevailingly shales with thin bedded sandstones while the 
upper portions are much more heavily bedded, the sandstones pre- 

••• • v . - 

Fig. 2 Spirifer mucron atus (Conrad), a characteristic and abundant index 
fossil of the Hamilton shales of the Cat.skiil margin 

vailing. The five divisions may possibly be more satisfactorily 
made on paleontologic characters than on physical, but in most of 
the advisory reports on economic and practical problems involving 
this district the subdivisions can not be emphasized. The wdiole 
scries is essentially conformable and is very little disturbed [see 
report on Milestone quarries, pt 2]. 

(10) Onondaga limestone. A bluish gray, massive, thick bedded 
cherty, somewhat crystalline limestone. It is strongly marked off 
from the Hamilton and Marcellus above, and, because of its greater 
resistance to erosion, usually forms a dip slope controlling stream 
adjustment and ultimately inducing the development of unsymmet- 
rical valleys w ith gentle easterly slopes and clifflike westerly borders 
where the streams are sapping the overlying Marcellus and Ham- 
ilton shales. It is not sharply separable from the Esopus below but 
everywhere in this region graduates into it with increase of silicious 



and argillaceous impurities. Estimating the formation from the drill 
cores that have penetrated it, and placing the lower limit as nearly 
as may be at the horizon of changes from predominant lime to pre- 
dominant silicious content, the approximate thickness in this region 
is placed at 200 feet. The rock where exposed exhibits considera- 
ble joint development and these are considerably enlarged by the 
solvent action of percolating waters. This factor is considered of 
some importance in connection with the other limestones of the 
district in aqueduct construction and permanence. The Onondaga 
has been used as a building stone formerly sold as marble, some 
grades of which are good stone. On the line of the aqueduct it is 
confined to the Rondout and Esopus valleys. The chief fossils are: 
Atrypa reticularis, Zaphrentis prolifica, 
Leptostrophia perplana, Platyceras dumosum, 
Leptaena rhomboidalis, Dalmanites selenurus. 

(11) Esopus and Schoharie shales (a slaty grit). The Schoharie 
as a distinct formation is not distinguishable in this region. The 
very thick and comparatively uniform, gritty, black, dense, almost 
structureless rock is a distinct unit. It is a silicious mud rock with 
very obscure sedimentation markings, but showing independent 
secondary cleavages induced by later dynamic factors, and, on long 
exposed surfaces always exhibiting chiplike fragments as the result 
of weathering. But it is not an easily destroyed rock. In so far 
as the bedding is obscure and the induced structure predominates, 
the rock is a slate; and in so far as it is distinctly gritty (sandy) 
instead of argillaceous it is a grit. The formation might therefore 
be more accurately designated as a slaty grit. The lack of plain 
bedding structure makes it impossible to estimate its thickness, 
since the foldings or other displacements can not be allowed for ; 
but the accumulated data of drill holes in more advantageous 
position indicate an approximate thickness of 800 feet. The rock 
is considered exceptionally good ground for the tunnel. 

A few fossils occur the most characteristic being Taonurus 
c a u d a g a 1 1 i . There are also in certain layers of limited extent, 
Leptocoelia acutiplicata and Atrypa spinosa. 

(12) Oriskany and Port Eivcn transition (silicious shaly lime- 
stone). There is no well defined and distinct separation here be- 
tween the Oriskany and the underlying Port Ewen, but because of 
the importance and persistence of the formation in other and re- 
lated areas the name is held. The equivalent of the Oriskany is in 
this district involved with a strongly developed transition zone 
which in physical features is intimately associated with the Port 



Ewen as a single unit. If any distinct formation is to be recog- 
nized it would be on the basis of transitional faunal character, 
placing the fossiliferous upper 100 feet in the Oriskany transition 
and confining the name Port Ewen to the rather unfossiliferous 
and concretionary, shaly, argillaceous limestone of the lower 100 

Fig. 3 Spirifer arenosus (Conrad), one of the characteristic index fossils of 
the Oriskany occurring in the Port Ewen-Oriskany transition 

This transition rock is strongly bedded, argillaceous and 
silicious limestone, very quartzose in certain layers, but there 
are no exposures in this area that would be called sandstones. 
Fossils are abundant and show marked Oriskany peculiarities. 
Those of most characteristic relations are : Hipparionyx 
proximus, Leptostrophia magnifica, Spirifer 
murchisoni, Spirifer arenosus, Platyceras 
nodosum, Strop hostylus expansus. 



(13) Port Ewen shaly limestone. The beds below those noted 
in the preceding paragraph are essentially argillaceous, shaly lime- 
stones. They vary from rather massive to thin bedded, are dark 
grayish in color, and have a peculiar nodular or concretionary de- 
velopment along certain sedimentation lines. These spots have less 
resistance to weather than the surrounding rock and therefore 
develop rows of pits along the face of an outcrop. Their size, 
6 to 18 inches or more across, together with their persistence makes 
an easily recognized physical feature. The few fossils that are 
found are not very characteristic. The following should be men- 
tioned : Spirif er perlamellosus. 

In the discussion and on the maps the Port Ewen and Oriskany 
are treated together as a single unit as the Oriskany-Port Ewen 

(14) Becraft limestone. Massive, heavy to thin bedded, light 
colored, semicrvstalline to thoroughly crystalline limestone. More 
massive beds very pure, 94 -1- i Cat ( ) ... Shaly beds resemble the 

New Scotland which they pass 
into at the base. The most char- 
acteristic features for field iden- 
tification are (a) pink or light 
colored spots, (£>) a more 
coarsclv crvstal'ine condition 
than any of the associated strata, 
(c) occasional large calcite 
cleavages to be seen wherever 
a fossil crinoid base A s p i d o - 

Fig. 4 Sieberella pseudogal- CrillUS SCUtelliformis 

e a t a Hall, the most characteristic index • i i / j\ .1 i 

fossil of the Beacroft limestone of the Ron- IS broken, (0 ) the Very CharaC- 

dout region • - . ,. , .. c . , . , 

tenstic fossil b 1 ebe rel 1 a 
pseudogaleata, and (e) many crinoid stems. 

The formation carries many, fossils in. addition to these given 
above, among which are Spirifer concinnus, Uncin- 
u 1 u s campbellanus. 

(15) A ezv Scotland slialy limestone. Thin bedded, dark gray to 
reddish sandy and shaly limestones. The rock breaks out in slabs 
on weathering and develops red iron stains. It has especially 
abundant fossils, the most characteristic of which are: Ortho- 
t h e t e s w o o 1 w o r t h a n u s , Spirifer macropleura. 
Other common ones are : L e p t a e n a r h o m b o i d a 1 i s , 
Strop honella headleyana, Ripidomella oblata, 
Strop heodonta becki. 



(16) Coeymans Hun-stone. Heavy bedded, dark gray, argillace- 
ous and flinty limestone. The characteristic features for field 
identification arc (a) abundant chert nodules, (b)- the occurrence of 
coral reef structure and heads of corals, Favosites h elder- 

Fig, s Spirifer macropleura (Conrad), the most characteristic index fossil 
of i he New Scotland beds in the Rondout region 

b e r g i a . The brachiopods S i e b e r e 1 1 a g a 1 e a t a and 
A t r y p a reticularis are very common. 

This formation has a thickness of about 8o feet and is rather 
distinctly separated from the underlying Manlius. The Coeymans 
is considered the base of the Devonic system of New York. It is 

Fig. 6 Sieberella galeata (Dalman). the most reliable index fossil of the 
Coeymans limestone of the Rondout region 

perfectly conformable upon the underlying series and it is evident 
that in this region there was no important break in the progress of 

c Siluric strata. (17) Manlius limestone. Lime mud rock, fine 
textured, dense, with plainly marked sedimentation lines, gray to 
dark gray color. The most characteristic features in the field are 
(a) fine texture, (£>) sedimentation lines, as if laid down in quiet 
waters as a lime mud, (c) solution joints sometimes enlarged to 



cavelike form into which surface streams disappear (such as Pom- 
pey's cave near High Falls), (d) mud crack surfaces (in lower 
beds), (e) occurrence of the fossil Leperditia alta. 

Its abundant jointing and the tendency to develop solution cav- 
ities from them is considered an objectionable character. 

(18) Cobleskill and cement beds (limestone). It is not pos- 
sible without the most painstaking, comparative, chemical and pale- 
ontologic research to differentiate the cement layers from the 
inclosing beds and to assign them all to the subdivisions that are 
recognized in some previous publications, 1 as the (a) Rondout 
cement (b) Cobleskill limestone, (c) Rosendale cement, and (d) 
Wilbur limestone. There are, however, two workable natudal ce- 
ment beds, both at Rondout and at Rosendale, with a nonworkable 
layer between each case, and also one between the lower 
and the next underlying formation. Whether the two cement beds 
at Rondout represent the Rondout and the Rosendale horizons 
with the Cobleskill between, or whether they should both be re- 
garded as Rondout with Cobleskill below, can not concern our 
present problems. And again, whether or not the two cement beds 
at Rondout are the same two that appear at Rosendale, or whether 
they are equivalent only to the upper one with a new lower bed 
(The Rosendale) added in this area and then with the Cobleskill 
between these two as claimed by Grabau, does not alter the plain 
fact that the whole series is a physical unit. It is a gray, rather 
close texture limestone, resembling the Manlius proper, and con- 
tains few fossils. It is perhaps even better yet to group all of 
these limestone beds below the Coeymans into a single unit and 
call it the Manlius series. 

(19) Binnezvater sandstone. Below the Manlius cement rock 
series lies the 60-100 foot Binnewater. It is chiefly a well bedded 
quartz sandstone, almost a quartzite in the upper beds with more 
shale in its lower portion, in color varying from white to greenish 
yellow and brown. The rock is rather porous in certain beds and 
especially along the bedding planes and is not well recemented 
where crushed by crustal movements. It is confined to the Rondout 

(20) High Falls shale. 2 Greenish to red argillaceous to sandy 
shales. The exposures are often a brilliant red while the rock 

1 N. Y. State Mus. Bui. 92 (Grabau). p. 311-13; N. Y. State Mus. Bui. 80 
(Hartnagel), p. 355-58; N. Y. State Mus. Bui. 69 (Van Ingen and Clark), 
p. 1 184, 1 185. 

2 The term given by Hartriagle. N. Y. State Mus. Bui. 80. p. 345. 



from drill cores is seldom highly colored. The protected beds are 
more commonly greenish in color and contain much iron sulphide. 
Occasional thin limestone beds occur in the upper portion at High 
falls — one of 4 feet forms the lip of the lower fall. The High 
Falls shale is confined to the Rondout valley and on the line of 
the aqueduct is 67-100 feet thick. 

(21) Shawangunk conglomerate. The Shawangunk is a con- 
glomerate and sandstone. The constituent pebbles are almost wholly 
quartz, well worn, and varying in size from that of sand to pebbles 
of several inches diameter. But for the most part the pebbles are 
small, abundantly mixed with sand, bound together by a silicious ce- 
ment. Rarely a true quartzite is developed and still more rarely a 
shaly facies. The rock is therefore very hard, brittle, and in the un- 
disturbed portions fairly impervious and resistant. But it suffers 
from crushing along zones of disturbance in folding and faulting 
and these zones are very imperfectly recemented. It is a durable 
rock, very resistant to ordinary decay, but forms great talus slopes. 
It is used for buhrstones (millstones), etc. It varies in thickness 
on the lines of the aqueduct from 280-400 feet. The rock is lim- 
ited in its northward extension to this district — southwestward 
it is much more broadly exposed in the continuation of the Shaw- 
angunk range. 

The Shawangunk completes the conformable Siluro-Devonic 
series down to the erosion interval at the close of the Ordovicic. 
The series of conglomerates, sandstones, limestones, and shales 
make an imposing column approximating 3000 feet of strata differ- 
entiated with more or less ease into 15 separate and mapable 
formations and a possible 5 or 6 more with careful paleontologic 
work. The series begins with the capping beds of the Shawangunk 
range and its northward extension toward the Hudson river at 
Rondout and Kingston, and thence westward constitutes the rock 
floor while its structures control the surface configurations far be- 
yond the limits of the region under consideration. Immediately 
to the north and partly within the area here treated is the famous 
Rosendale cement region, the pioneer cement district of America 
and for many years the best producer. The strata used 
are almost exclusively the upper members of the Siluric 
{" cement beds ") closely associated with the Cobleskill between 
the Manlius proper and the Binnewater sandstone. Rarely the Be- 
craft from the Devonic series furnishes some cement rock. 

/ Cambro-Ordovicic formations. Between the Precambric 
metamorphics of the Highlands beneath and the Siluro-Devonic 

4 6 


sediments of the Shawangunk range and the Catskills above, lies 
a series of quartzites, limestones and slates less complexly dis- 
turbed than the older and more disturbed than the younger series 
— set off from both by unconformities representing time intervals 
that cover both folding and erosion. They are of more than 4000 
feet thickness — how much more it is impossible to estimate be- 
cause of the obscurity of data in the slates. There are very few 
fossil forms preserved in them. The series is, however, readily 
and sharply separable into three formations that may be mapped 
upon lithologic characters alone. They are of most importance in 
the Wallkill valley, Moodna creek, Newburgh, Fishkill, New Ham- 
burg and Poughkeepsie districts. Their character, structure, and 
conditions have required careful consideration in the decisions on 
the Wallkill and Moodna siphons and in the discussions on the 
proposed Hudson river crossings [sec Hudson river crossings, 
pt 2]. 

(22) Hudson River slates. The upper member of the Cambro- 
Ordovicic series is in itself complex. Prevailingly it is a slaty 
shale, occasionally it is a sandstone or shaly sandstone, or a simple 
shale ; still more rarely it is almost a true slate, and very rarely 
a phyllite. The constituents vary from prevailing clay to quartz 
sand repeatedly in almost every locality. It is probable that as a 
rule the upper portions are the more heavily bedded and arena- 
ceous. The rock is excessively affected by the dynamic movements 
that have at least twice disturbed it. A slaty cleavage in the more 
argillaceous members is most noticeable, but almost everywhere the 
strata are strongly tilted, crumpled, broken, faulted, or crushed in a 
most confusing way. This together with an original obscurity 
in bedding, and the obliteration by subsequent shearing of much 
that did exist, makes it impossible to reconstruct the complicated 
structure or compute the thickness of the formation. It is of such 
physical character as to absorb within its own limits much of the 
disturbing movements, and neither the formations above nor imme- 
diately below are so extensively and intimately affected. The 
formation is widely exposed and forms the bed rock over very large 
areas. Almost everywhere it is impervious to water, easy to pene- 
trate by drill or tunnel, and resistant to decay. A few Ordovicic 
fossils may be found, the most characteristic being D a 1 m a n e 1 1 a 

(23) Wappingcr limestone} (In part Cambric, and in part 

1 The Wappinger Valley limestone of Dwight (1879) and Dana. The 
Wappinger limestone of Darton and others. 



Ordovicic). The formation is prevailingly of a compact, fine 
texture, dark gray, either massive or strongly bedded limestone. 
Where the stratification is very plain there are light and dark layers 
and an abundant silicious intermixture. In many outcrops the rock 
is so massive that even the dip and strike are obscure. Some places 
the rock is fine crystalline, almost a micromarble. On weathered 
surfaces it almost always exhibits a crisscross etching which marks 
the traces of rehealed cracks. From these it is seen that many of 
the apparently massive compact beds have at one time been exten- 
sively crushed. In many places there is scarcely a square inch 
wholly free from these evidences. The formation is best exposed 
in the wide belt that extends southwestward from the vicinity of 
Poughkeepsie and crosses the Hudson at New Hamburg into the 
Newburgh district. It undoubtedly underlies the slates in the rest 
of the adjacent area. There are few fossils and they are rarely 

(24) Poughquag quartsite. ■ Below the Wappinger limestone and 
upon the upturned and eroded edges of the Highlands gneisses lies 
a quartzite of variable thickness but which reaches at least 600 feet. 
It is a strongly silicified quartz sandstone — a quartzite by indura- 
tion. It is strongly bedded but seldom shaly. Traces of schistosity 
may appear in certain zones and this is somewhat strongly developed 
outside of the area at the type locality (Poughquag, N. Y.). 

Only fragments of trilobite spines have been found in this forma- 
tion within the district. 

g Later crystallines south of the Highlands. South of the 
Highlands proper except at one locality (Peekskill creek valley and 
its southwestward continuation through Tompkins Cove and Stony 
Point) the rocks are all much more thoroughly crystalline. There 
are two formations, and in places traces of a third, above the Gren- 
ville gneisses (Fordham gneisses and associates). These are known 
locally as Manhattan schist, Inwood limestone, and Lozvcrrc quartz- 
ite. In Westchester and New York counties the quartzite 
is rarely found, and in a considerable proportion of those places 
where it does occur its relations are more consistent with the 
gneisses below than with the limestone-schist series above. This 
is true indeed of the type locality (Lowerre). There are, however, 
at least two points where the occurrence favors the reverse inter- 
pretation, so far as any is shown, and therefore a quartzite may be 
regarded as finishing the series, and making uncertain but probably 
unconformable contact with the underlying gneisses. 



This series together with the gneisses below constitutes the bed 
rock and controls the underground conditions for all of the line 
south of the Moodna valley, 50 miles above New York. All of the 
southern aqueduct, and the New York city distribution conduits are 
wholly concerned with these rocks, and two divisions of the northern 
aqueduct have a large proportion of their work in them. 

It is not wholly clear what age these crystallines represent. It 
is certain that the underlying gneisses are Grenville and that the 
metamorphic quartzite, Inwood, Manhattan series, is Post-gren- 
ville. It is possible that these latter are also Precambric. But 
usage following the correlations of Dana 1 and in the absence of 
as good evidence from any other source has regarded them as the 
Cambro-Ordovicic crystalline equivalents of the Poughquag-YYap- 
pinger-Hudson River series of the north side of the Highlands. 
The writer has elsewhere shown 2 that the evidence and arguments 
are not all on one side and that considerable doubt may still be 
entertained on that point. There is no object in following that 
argument here or in modifying the treatment here followed of mak- 
ing them a distinct series. Even if they should prove to be the 
exact equivalents of the Hudson River- Wappinger-Poughquag 
series the formations are physically so different and require so 
different treatment in discussion that they must for our present 
purpose be regarded as an essentially distinct series. From that 
standpoint alone the usage here followed is justified. The Man- 
hattan schist of Westchester county as a type differs as much 
petrographically from the Hudson River formation of the New- 
burgh district as the Catskill formation of Slide mountain differs 
from the Jameco gravels of Long Island. In a discussion where 
physical or petrographic character is in control there is no doubt 
about the advisability of treating the two separately. 

(1) Manhattan schist* This is primarily a recrystallized sedi- 
ment of silicious type. It occurs as a nearly black or streaked, 
micaceous, coarsely crystalline, strongly foliated rock. The chief 
constituents are biotite, muscovite and quartz. Quartz, feldspar, 

1 Dana, J. D. On the Geological Relations of the Limestone belts of 
Westchester county, N. Y. Am. Jour. Sci. 20:21-32, 194-220, 359-75, 450-56 
(1880); 21:425-43; 22:103-19, 313-15. 327-35 (1881). 

2 Berkey, Charles P. " Structural and Stratigraphic Features of the 
Basal Gneisses of the Highlands." N. Y. State Mus. Bui. 107 (1907), 
p. 361-78. 

3 Manhattan schist of Merrill. N. Y. State Mus. 50th An. Rep't, 1:287. 
Same as " Hudson schist," of N. Y. city folio no. 83. 



garnet, fibrolite and epidote also occur in large quantity. Occa- 
sional streaks or masses are hornblendic instead of micaceous. 
These are interpreted as igneous injections. They are especially 
abundant on Croton lake and near White Plains. 

It is essentially a quartz-mica schist. But it is almost everywhere 
very coarse textured and hardly ever exhibits the fine grained, uni- 
form structure of typical schist. Its abnormal make-up — the pre- 
dominance of biotite and quartz — is the best defense for its petro- 
graphic classification. The abundance of mica makes it a tough rock 
but not very hard. The joints and fractures formed in later move- 
ments are not healed and zones of bad shattering are susceptible to 
considerable decay. These crushings are sufficiently common to en- 
courage borings to tap their content of water for small family use 
throughout Westchester county ; but they do not represent large 
circulation in any case. On the whole, the rock if fresh is good 
and durable. It may, though rarely, carry considerable sulphide. 
Practically all of the strictly original sedimentation marks are de- 
stroyed by metamorphism. The formation has great thickness, but 
because of the destruction of original bedding lines by recrystalli- 
zation and additional complication by most complex folding, shear- 
ing, crushing and faulting, the structure can not fully be unraveled 
and the thickness can not be estimated with any approach to ac- 
curacy of detail. But there is probably a thickness represented of 
several thousand feet. 

(2) Inwood limestone or dolomite. This formation lies beneath 
the Manhattan. It is everywhere coarsely crystalline either massive 
or strongly bedded, often very impure with development of second- 
ary (recrystallized) mica (phlogopite) and other silicates, espe- 
cially tremolite. It is essentially a magnesian limestone or dolomite 
in composition. There is an occasional quartzose bed in the midst 
of the limestone as at East View. The upper beds are most charged 
with mica and occasionally beds attacked by alteration have much 
green, flaky chlorite. There are occasional interbeddings of lime- 
stone and schist as a transition fades. 

The coarser grades upon exposure to weathering readily yield by 
disintegration to a lime (calcite) sand resembling rouehlv an ordi- 
nary sand in general appearance. At Inwood, the type locality, this 
disintegration is so pronounced that great quantities are readily 
shoveled up and used for various structural purposes in the place 
of other sand. This dolomite is especially liable, as now shown by 
extensive explorations, to serious decay to great depth. The under- 
ground circulation seems to attack the micaceous beds with great 


success and in some places the residue after this solvent action is 
of the consistency of mud. A nearly vertical attitude of the beds 
accentuates the opportunity. The most troublesome piece of ground 
encountered on the whole line of the New Croton aqueduct, con- 
structed in 1885, was in a weak zone and crevice in the Inwood near 
the village of Woodland on the margin of the Sawmill valley [see 
discussions of Bryn Mawr siphon and New York city distributions 
in part 2]. 

The thickness probably varies but in many places where there is 
only a narrow limestone belt it is due more to shearing or faulting 
out than to original thinning. The most satisfactory estimates are 
based on the explorations at Kensico dam and the field observations 
at I52d street. They indicate- an approximate thickness of 700 feet. 
But in all cases either the margins are obscured or there is possibility 
of faulting to modify measurements. There are no fossils. Weath- 
ering and erosion has almost everywhere developed valleys or de- 
pressions especially small tributary valleys in all formations, but as 
pointed out years ago by Professor Dana the principal valleys pre- 
vailingly coincide with the limestone belts. 

(3) Lowcrre quartzite. At Hastings-on-Hudson and again 
near Croton lake, there is a quartzite that appears to be 
conformable with the Inwood above. There is possibly more than 
50 feet. It is a simple, clean quartzite. The other quartzites of 
Westchester and New York county have a more distinct relation- 
ship to the underlying gneisses with which they are conformable. 
The Lowerre of the type locality is of this second class. In the 
great majority of places where this bed would be expected to occur 
there is not a trace of it. 

h Older metamorphic crystallines (Grenville series). 1 "The 
lowest and oldest, as well as the most complex in structure and rock 
variety, of all the formations of the Highlands region of south- 
eastern New York is essentially a series of gneisses." Cutting 
these gneisses as intrusions of various forms are a great number 
and variety of more or less distinctly igneous types. In form they 
vary from small dikes or stringers to' great batholithic masses; in 
composition, from the extremely basic peridotites or pyroxinites of 

1 This interpretation of the larger relations of the complex gneisses 
constituting the basis of the series, lying below the Manhattan-Inwood- 
Lowerre series, was presented by the writer under the title: Structural 
and Stratigraphic Features of the Basal Gneisses of the Highlands. N. Y. 
State Mus. Bui. 107 (1907). p. 361-78. The accompanying description is 
largely an abstract of this paper. 



the Cortlandt series to the very acid granites of Storm King moun- 
tain or the granophyric pegmatites of North White Plains; and in 
relative age they likewise vary from a period antedating the chief 
early metamorphic transformation of the Grenville to Postman- 
hattan time. Put these clearly igneous types attain a considerable 
prominence as separable units in the practical consideration of the 
problems of the project and on that account the chief ones will be 
more fully described under the next group. 

The older portion — the various schists, banded gneisses, quartz- 
ites, quartzose gneisses, graphitic schists, and serpentinous and 
tremolitic limestone, forming the complex through which and into 
which the igneous masses have been injected — form together an 
interbedded series that was originally a sedimentary group. There 
is nothing known that is older in this region. Us characteristics and 
relations mark it as in all probability the equivalent of the " Gren- 
ville " of the Adirondacks and Canada. 

No single type and no single characteristic can be given as a 
simple guide to the identification of this formation. The prevalence 
of certain varieties or groups of these and the strongly banded 
structure give a certain degree of character that forms a reason- 
able working base. The formation includes banded granitic, horn- 
blendic, micaceous and quartzose gneisses ; mica, hornblende, 
chlorite, quartz and epidote schists; garnetiferous, pyritiferous, 
graphitic, pyroxenic, tremolitic, and magnetitic schists and gneisses ; 
crystalline, tremolitic, and serpentinous limestones, aphi-dolomites, 
serpentines and quartzites ; pyrite, pyrohitite and magnetite de- 
posits. This is the basal series. Put it is complicated by a multi- 
tude of bands of granitic and dioritic gneisses that represent 
injections of igneous material at a time sufficiently remote to be 
subjected to most of the early metamorphic modifications. The 
equally abundant occurrences of quartz stringers and pegmatite 
lenses though of later origin can not be separated from this com- 
plex mass and the whole must be regarded as a physical unit. The 
occurrence of interbedded limestones and quartzites together with 
a variety of conformable schists and banded rocks, marks the 
formation as essentially an old recrystallized sediment. 

No member of this older unit of the basal complex is sufficiently 
prominent to indicate a great break cr change up to the time of 
the first great dynamic movements and igneous outbreaks. The 
following comparatively constant members are sometimes persistent 
enough to be considered formational units, but even more commonly 



are obscure as to boundaries or are of too small development to 
map separately. 

(4) Interbeddcd quartzite. Always a quartzite schist and 
always exhibiting conformity with the banded gneisses and schists. 
This is regarded as the uppermost member. 

(5) Fordham gneiss (Banded gneiss). Granitic and quartzose 
black and white banded gneisses and schists of very complex com- 
position and structure. 

(6) Interbeddcd limestones. Crystalline. Interbedded, very 
impure, serpentinous and tremolitic. granular dolomites, usually 2 
to 50 feet thick, possibly reaching a thickness of more than 100 
feet in a few cases. 

(7) Older intrusive gneisses. Variable types, mostly granites or 
diorites, strongly foliated sills. 

Many are of very obscure relations. The line of close distinction 
between recrystallized sediment, segregations accompanying that 
change, and true igneous injection can not be drawn. 

i Special additional igneous types. Under this heading are 
included the massive or little modified, not at all or only moderately 
foliated, igneous masses of later origin and local rather than re- 
gional development. In some cases, however, they are of decidedly 
controlling importance in the local geology and rise to the status 
of definite formations. The most noteworthy of these within reach 
of the aqueduct explorations are : 

(8) The Storm King Mountain gneissoid granite 

(9) The Cat Hill gneissoid granite (central Highlands) 

(10) The Cortlandt series of gabbro-diorites (near Peekskill) 

(11) The Peekskill granite (east of Peekskill) 

(12) The Ravenswood granodiorite (Long Island City) 

(13) The pegmatite dikes and lenses (segregational aqueo- 

igneous type) 

(8) The Storm King gneissoid granite is one of the largest of 
the clearly igneous and less completely foliated types. It consti- 
tutes the whole of Storm King mountain and the larger part of 
Crows Nest on the west side of the Hudson, and, crossing the river, 
forms the chief rock of Bull hill and Breakneck ridge. It is a 
rather acid, coarse grained, reddish granite with considerable 
gneissoid structure in a large way [see Hudson river crossings, 

(9) The Cat Hill gneissoid granite is not essentially different 
from the Storm King type as a physical unit. Its occurrence at a 



different point (Cat hill), widely separated by other types from 
the Storm King locality, and in rather large development, is worthy 
of separate note. It is cut, of course, in the long tunnel through 
Cat hill. 

(10) The Cortlandt series of gabbro-diorites occupies an area 
of about 20 square miles between Peekskill and the Croton river, 
nearly all on the east side of the Hudson. It includes a very com- 
plete range of coarse grained, massive, igneous rocks from soda 
granites, grano-diorites and quartz-diorites to true diorites, norites, 
gabbros, pyroxenites, and peridotitcs. They doubtless represent 
stages or portions in the differentiation of a magma. The inter- 
relations are only partially determinable, and the petrographic dis- 
tinctions in detail are not useful here. The area occupied by the 
Cortlandt series has an uneven hilly surface with no structural 
trend, and makes the most striking contrast to the ridge and longi- 
tudinal valley structure of the rest of the region of the crystallines. 

(11) The Peekskill granite, a white, or pink massive, very coarse 
grained, soda granite, occupying approximately 4 square miles im- 
mediately north of the Cortlandt area 2 miles east of Peekskill, 
is believed to be genetically related to the Cortlandt series. The 
evidence in favor of such a relationship has been gathered in the 
prosecution of this work and has not been published. But it may 
be said that the textures, structure, age, relationship to older crys- 
tallines, interrelations with the Cortlandt series, consanguinity of 
mineralogy, and composition all point toward the above relation- 
ship. In essential relations, therefore, it is the acid extreme of the 
Cortlandt series. Its economic features, however, are of sufficient 
importance and its easy differentiation from the regular Cortlandt 
types require that it should have separate treatment. 

(12) The Rravenszvood grano-diorite occurs chiefly in Brooklyn. 
It is a slightly foliated mass intrusive in the Fordham gneiss and 
is doubtless connected in origin with the sources of many of the 
hornblendic intrusive bands in the Fordham and Manhattan forma- 
tions in the district. It covers a known area of about 5 or 6 square 
miles and may be more extensive. The rock is suitable for struc- 
tural material and has required consideration in the study of " Dis- 
tributary conduits " [see pt 2 East River section]. 

(13) Pegmatites. The pegmatites and pegmatitic granophyric 
masses of all kinds are of almost universal distribution in the 
foliated crystallines. They vary from quartz bunches or stringers 
to pegmatitic lenses and irregular masses, and to definite granitic 



or pegmatic dikes. In many places they constitute a large propor- 
tion of the formation in which they occur. They doubtless vary 
in age, but for the most part seem to belong to the later period of 
metamorphism. Many of them are massive and largely free from 
foliation. They no doubt have a complex origin between simple 
aqueous segregation on the one side and true igneous intrusion on 
the other. 

Summary of formations 

Group a Quaternary deposits 

(i) Glacial drift -v Occurs as a surface 

Till and modified drift, extra mantle over nearly all 
marginal outwash, sands and > of the region under 
gravels, etc. discussion, except the 

immediate sea margin 


Group b Tertiary and Cretaceous deposits 

(2) Tertiary outliers 

(a) Pliocene littoral deposits 


(b) Miocene " fluffy " sand (Beacon 


(3) Upper Cretaceous beds 

(a) Lignitiferous sand (marl series) 

(b) Matawan beds (clay marls) 

(c) Raritan (clays and sands) 

Confined to Long Is- 
land, Staten Island 
and the New Jersey 



Group c Jura-Trias formations 
(4) Palisade diabase intrusion 

(5) Newark series of conglomerates, 
sandstones and shales 

Confined to the west 
side of the Hudson 
south of the High- 





Group d Dcvonic strata 

(6) Catskill, white and red conglom- 

erate (1725 feet) 

(7) Oneonta (upper flagstone) (3000 


(8) Ithaca and Sherburne (lower flag- 

stone) (500 feet) 

(9) Hamilton and Marcellus shales 

(flagstone and shales) (7°° 

(10) Onondaga limestone (200 feet) 

(11) Esopus and Schoharie shales 

(silicious) (800 feet) 

(12) Oriskany and Port Ewen transi- 

tion (100 feet) 

(13) Port Ewen limestone and shale 

(150 feet) 

(14) Becraft limestone (75 feet) 

(15) New Scotland shaly limestone 

(100 feet) 

(16) Coeymans cherty limestone (75 

Confined to the Cats- 
kills, the Esopus and 
Rondout valleys, the 
northern extension of 
the S h a w a n g u n k 
range, and Skunne- 
munk mountain near 


(17) Manlius limestone (70 feet) 

(18) Cobleskill limestone and cement 

beds (30 feet) 

(19) Binnewater sandstone (50 feet) 

(20) High Falls shale, including small 

limestone beds (75-80 feet) 

(21) Shawangunk conglomerate (250- 

350 feet) 

Siluric strata 

^ Confined to the Rondout 
and Esopus valleys 
and the northerly ex- 
tension of the Shaw- 
a n g u n k range, 
through the cement 
region of Rosendale, 
Binnewater, Rondout 
and Kingston, and a 
small outlier at Skun- 
nemunk mountain 




Group f Cambro-Ordovicic formations 
(22) Hudson River slates, shales, and Especially prominent as 

sandstones (very thick) (Or- 
dovicic) more than 2000 feet 
(23) Wappinger limestone (1000 feet) 

( in part Cambric and in part I valley, and the region 

surface formations in 
the Shawangunk 
range, the Wallkill 

(24) Poughquag quartzite (600 feet) 

eastward and south- 
ward to the High- 
lands, on both sides 
of the Hudson 

Group g Later crystallines (South of the Highlands) 

(1) The Manhattan 

(Uncertain age 
schist, a thor- 

oughly and coarsely crystalline 
sediment of uncertain age — 
generally supposed to be equiva- 
lent to the Hudson River slates, 
(Ordovicic) but here separated 
without necessarily raising that 
question because of their very 
different physical and petro- 
graphic character 

(2) Inwood limestone (or dolomite), 

a magnesian crystalline lime- 
stone of uncertain age, generally 
supposed to be the equivalent of 
the Wappinger (Cambro-Ordo- 
vicic), but here enumerated sep- 
arately without necessarily rais- 
ing that question because of 
their very different lithologic 
character and associates 

(3) Lowerre quartzite, an occasional 

quartzite of uncertain relations 
and very limited development 

Confined to the region 
east of the Hudson 
river and south of 
the Highlands proper, 
occupying the region 
from the Highlands 
to Long Island 




Group h Older crystallines (Highlands gneisses) 

(Grenville series of metamorphics and intrusives — Precambric) 
(4) Interbedded quartzite 



quartzose schist 

(5) Fordham gneiss (chiefly 

sedimentary). Granitic 
and quartzose banded 
gneisses and schists of 
very complex develop- 

(6) Interbedded 

with the 

(7) Old intrusions 

ble masses of granitic gneisses 
of igneous origin cutting the 
Grenville series, such as Storm 
King granite, Cat Hill granite, 



Large and varia- 


istic of the 
lands and some 
larger ridges extend- 
ing southward to 
New York city. A 
series, which in petro- 
graphic variety, is as 
complex as all of the 
rest of the forma- 
tions of the region 

Postgrenville in age 

Group i Special additional 

(8) Storm King gneissoid granite, 

Storm King-Breakneck district 

(9) Cat Hill gneissoid granite. Garri- 

son district 

(10) Cortlandt series of gabbro-diorites. 

Peekskill-Croton district 

(11) Peekskill granite. A boss, related 

to the Cortlandt series. Peeks- 
kill district 

(12) Ravenswood grano-diorite. A 

boss. Brooklyn, Long Island 
City and Southern Manhattan 

(13) Pegmatites. Dikes, lenses, segre- 

gations of general distribution 

igneous types 

~\ These are masses of 
strictly igneous origin 
(except the pegma- 
tite) and of larger 
development which 
either because of 
their abundance (peg- 
matites) or large area 
(Cortlandt) or eco- 
nomic features 
(Peekskill) or im- 
portant bearing upon 
the plans of the aque- 
duct (Storm King) 
are worthy of sepa- 
rate note. 



3 Major structural features 

In addition to the simpler structural characters of the strata, 
already sufficiently emphasized in the individual descriptions, there 
are numerous others of more general relation whose value and in- 
fluence it is necessary to consider in many of the practical problems. 
Those of most importance are the unconformities, folds and faults. 
They are directly related to continental elevation and subsidence, 
to mountain forming movements and denudation processes, to meta- 
morphism and to igneous intrusion. 

a Sedimentation structures. In the younger strata the prin- 
cipal structures are those of bedding, stratification, conformable 
-succession, etc., characteristic of all sediments of such variety of 
type. These are prominent in the older groups of formations down 
to the crystallines, but the earlier Paleozoics are also affected sc 
profoundly by folding and faulting that attention is more concerned 
with these induced or secondary structures. 

Unconformities. Time breaks, with more or less disturb- 
ance of strata and accompanied by erosion, are numerous. 

(1) That between the glacial drift and the rock floor is the most 
profound. It causes the glacial drift to lie in contact with every 
formation of the region from the oldest gneisses of the Grenville 
series of the Highlands to the traces of Miocene beds of Long 

(2) The interval between the Pliocene and the Upper Creta- 
ceous beds is more obscure and hardly reaches the importance of 
an unconformity. It is probably more nearly of the value of a 
disconformity or of an overlap, and the very limited development 
of the overlying beds in the region gives little chance for determin- 
ing relations in much detail. 

(3) The overlap and unconformity between the Cretaceous and 
Triassic. A condition determinable only on the New Jersey side 
of the Hudson river. 

(4) The unconformity between the Triassic and underlying 
formations of different ages. An interval representing mountain 
development and extensive erosion* in which the chief movement 
probably belongs to the close of Paleozoic time and includes the 
Appalachian folding. 

(5) Unconformity between Siluric and the Ordovicic strata. An 
interval representing mountain development, folding and erosion, 
in which the movement known as the Green Mountain folding took 



(6) Unconformity between the Poughquag (Cambric) quartzite 
and the underlying crystallines. An interval in all observable cases 
of great length and profound changes involving mountain folding, 
metamorphism of the profoundest sort, and extensive erosion. 

(7) Among the crystallines of the south side of the Highlands 
there is one break of similar importance, between the Inwood lime- 
stone and the underlying gneisses. Whether or not it is the same 
as no. 7 above is not clear, but even if it represents the same break 
the relations are somewhat different in degree and character because 
of the lack of quartzite in almost all cases. 

Within the gneisses of the Grenville series and their associates 
of all kinds there are no breaks of the unconformity type known. 
The contacts are eruptive in character, or are displacements 

c Folds and mountain-forming movements. All of the forma- 
tions from the oldest up to and including the Lower Devonic strata 
are folded. Many of the smaller (minor) folds exhibit complete 
form in the stream gorges of the district, but all of the larger ones, 
the main folds, have in earlier time been eroded to such extent 
that the series is beveled off and only the truncated edges are to 
be seen, exhibiting strata standing more or less perfectly on edge, 
and making restoration of the form a very difficult or impossible 
task. This is only partially accomplished in the Siluro-Devonic 
margin along the Shawangunk range ; it is more complete in the 
Cambro-Ordovicic north of the Highlands, and it reaches its most 
perfect development in the crystallines of the Highlands and New 
York and Westchester counties. These differences correspond 
roughly to the differences in age of the strata, and, taken together 
with the evidence of the profound unconformities, indicate that 
mountain-forming movements of far-reaching importance visited 
the region no less than three times. Each time of such disturbance, 
of course, the underlying older series was affected by the move- 
ments of that epoch in addition to any previous ones, and as a con- 
sequence the older is to be expected to show more complexity of 
such structures. Each succeeding series separated by such activity 
is therefore one degree simpler in structure. 

Of these three epochs of great disturbance, one is (1) Precambric 
and corresponds to the time interval marked by the unconformity 
between the Poughquag quartzite and the gneisses; a second (2) is 
Postordovicic and corresponds to the time interval marked by the 
unconformity between the Hudson River slates and the Shawan- 



gunk conglomerates, and the last (3) is Postdevonic (probably 
Postcarbonic, judging from neighboring regions of similar history) 
and has left as its most important evidence in this district, the 
excessively complicated sharp foldings and thrusts of the Shawan- 
gunk range and its extension in the Rosendale cement district. 

Kinds. As to forms produced there are no usually described 
types that are not to be found here. The simpler forms of anti- 
clines and synclines, both open and closed, symmetrical and unsym- 
mctrical and overturned, are all common. The isoclinal is common 
in the gneisses. In each epoch of folding the compression forces 
were effective chiefly in a northwest-southeast direction producing 
arches and troughs whose axes trend northeast-southwest. This i.s 
the trend of the main structures throughout the region. 

The extent of crustal shortening accomplished by this series of 
compressions is undetermined, but that it amounts to a total of 
many miles is indicated by the fact that over broad areas the 
strata stand almost on edge. Furthermore, in the older Highlands 
and in portions of the Hudson river districts the folds have been 
slightly overturned so that commonly the strata on both limbs dip 
in the same direction (toward the southeast). This seems to indi- 
cate a strong thrust from the southeast. All stages between the 
gentlest warping to strongly overturned folds, and from minute 
crumbling to folds of great extent and persistence are to be seen. 

The effect of all the folding is chiefly to present a series of up- 
turned strata to erosion and encourage a subsequent development 
of valleys along the softer beds bordered by ridges of the more 
resistant types. 

As the axes of the folds lie in a northeast-southwest direction, 
this gives a marked physiographic development of ridges and val- 
leys of the same trend, a most conspicuous topographic feature of 
southeastern New York. 

d Faults. Accompanying the folding in each epoch, and 
especially the stronger overthrust movements there has been a 
tendency to rupture and displacement. These breaks are known 
as faults. Multitudes of them are of minute proportions and prac- 
tically neglectable in a broad view, but many also are of large 
extent, traceable across country for many miles and indicating dis- 
placements in some cases of many hundreds of feet. For the most 
part these faults are of the thrust type and wholly consistent with 
the folds in origin. They run generally in a northeast-southwest 
direction, especially the larger ones, and frequently form the sep- 
aration planes between different formations. Occasional cross 



faults occur (with northwest-southeast direction across the strike), 
but so far as is known they are always of minor consequence. In 
rare instances, the trace of a fault line on the surface describes 
curious curves, such as that at Cronomer hill above Newburgh, 
apparently inconsistent with the chief structural trend, but a study 
of the whole geologic relation in such cases shows them to be con- 
nected with the projecting spurs of underlying formations which 
in any large thrust movement plow their way with some success 
through the younger overlying, less resistant, strata. They differ 
in no material way from the ether more simple looking lines. 

Both normal and thrust faults occur, but the thrust type appears 
to be most common. 

The amount of displacement or throw is extremely variable. The 
larger faults represent movements of several hundred feet. In 
rare cases the movement may be as much as 2000 feet. 

The effects may be grouped as follows: (1) the appearance of 
formations out of their normal order, i. e. contacts between forma- 
tions that do not normally lie next to each other; (2) the produc- 
tion of escarpments, i. e. steep cliff-bordered ridges; (3) the de- 
velopment of zones of more or less extensively crushed rock along 
the principal plane of movement; (4) the determination of loca- 
tion for stream courses and gulches and valleys that cross the 

All of these effects are more noticeable and better preserved for 
the later movements than for the earlier ones. Many of those 
dating back to the earliest epoch, affecting only the crystalline rocks 
of the Highlands, are not readily detected. Most of the breaks 
have been healed by recrystallization and the contacts are often 
as close and sound as any other part of the formation. 

But this is not so true of the later epochs — and in them a good 
deal depends upon the type of rock affected. The more brittle and 
hard and insoluble types are more likely to still have open seams 
and unhealed fractures than the softer and more easily molded 
formations. In some of these, recent water circulation has still 
further injured the fault zones by introducing rock decay to con- 
siderable depth. Because of the more ready circulation in them, it 
is noticeable that some of the extensive decay effects are produced 
in crystalline rocks that otherwise very successfully resist destruc- 
tion. On the whole the softer clay shales and slates are less likely 
to preserve open water channels of this sort than any other forma- 
tion of the region. 



No part of the region is wholly free from faulting effects, except 
perhaps a part of Long Island. The Catskills also are very little 
affected — so little that this type of structure has not require con- 
sideration in the vicinity of Ashokan reservoir. But all parts of 
both the northern and southern aqueduct system have had this 
feature to consider. 

Further discussion of the specific local prohlems introduced by 
faulting and folding is given under the problems of part 2. A con- 
siderably more extended comment on the age of fault movement 
is given under the heading " Postglacial faulting." 

4 Outline of geologic history 

Most of the genera! features of geologic history have been 
involved more or less in the foregoing discussion. It is im- 
possible to wholly separate matters that are so intimately inter- 
related even though it is convenient to think of or consider one 
phase at a time. But it may serve a useful purpose to summarize 
the steps of progress as illustrated by local geology from the earliest 
geologic time to the present. 

a Earliest time. (Prepaleozoic, Agnotozoic, Proterozoic, or 
Azoic Era). There is little doubt that the oldest rocks known in 
this region are representatives of a time of regular sedimentation. 
Conditions favored the deposition of silicious detritus of variable 
composition with an occasional deposition of lime, nearly always in 
very thin beds. What these sediments were laid down upon or 
where they came from are unsolved questions. The remnants of 
them that are still preserved are the basis of the " Grenville series " 
as interpreted in this area, and are the basal (oldest) members of 
the " Fordham " or " Highlands gneisses." 

How long ago this series was deposited is not known. It can be 
stated only approximately even in the rather flexible terms used in 
historical geology. It is older than any Paleozoic strata (Pre- 
cambric), probably very much older. It is even possible that this 
series is as much older than the Cambric as that period is compared 
to the present. In short, it is not known, and there is apparently 
little immediate likelihood of finding out even to which of the sev- 
eral subdivisions of the Prepaleozoic this series belongs. It is cer- 
tain that before the Cambric sandstones of the Paleozoic era had 
begun to form, this older series was disturbed by crustal move- 
ments, folded, metamorphosed, intruded by igneous injections, ele- 
vated above the water (sea) level of that time and eroded by sur- 
face agencies. These movements and steps there is no doubt of. 



When subsidence 1 again depressed the area beneath the sea the 
deposition of sands that we now call Cambric (Poughquag) quartz- 
ite began. 

/; Early Paleozoic time. With the sedimentation upon this old 
crystalline rock floor a long time of apparently continuous deposi- 
tion began which ultimately resulted in the accumulation of several 
thousand feet of sandstones, limestones, and sandy or clayey shales 
that are now known as the Cambro-Ordovicic series (Poughquag- 
Wappinger-Hudson River series). But at the close of Ordovicic 
time or late in that period another crustal revolution began. The 
whole region was again compressed into mountain folds, faulted, 
sheared, metamorphosed, elevated above sea level, and subjected to 
erosion. This corresponds to the Green mountains folding of 

With the next subsidence and a return of sedimentation a new 
series began to form. The break marking the occurrence of all 
these changes, known locally as the Postordovicic unconformity, 
represents a considerable portion of Siluric time. 

c Middle Paleozoic time. The earliest deposits of this series, 
which continued to accumulate through late Siluric and all of De- 
vonic time, were heavy conglomerates very unevenly distributed 
over the new rock floor. These are the so called Shawangunk con- 
glomerates, a formation that within the boundaries of this imme- 
diate area and within a distance of 20 miles varies from a thickness 
of more than 300 feet to almost nothing. But for the most part, 
sedimentation was regular and fairly continuous and of immense 
volume. The whole series of conglomerates, sandstones, shales, 
grits and limestones belonging to the later Siluric and the Devonic 
are included. Not all are believed to be marine however. The 
Catskill and Shawangunk conglomerates may well be of continental 

Long after the deposition of all of these strata another crustal 
disturbance, for at least the third time, repeated the process of 
mountain-folding and erosion. This was the time of the Appalach- 
ian mountain-folding. In this region it caused a wonderfully com- 
plex development of folds and faults that are especially important 
and determinable as to type and age in the Rondout cement region. 
The movement, of course, affected all of the older formations as 

1 There may possibly be an intermediate stage, practically a duplication 
of the whole as given above, between the very oldest and the Cambric, 
represented in the " later crystallines," but this may as well be neglected 
for the present. 



well, but on them, already disturbed by earlier displacements, the 
features chargeable to the disturbance can not always be distin- 
guished from older ones. All three of the mountain-forming com- 
pressions seem to have been controlled by the same relationship 
of forces and adjustments of movement, for the results are in each 
case the production of folds or faults of similar orientation and a 
final structure of uniform trend. 

Deposition had been going on for ages, chiefly on the west and 
north side of the older crystallines ; but with a return of sedimenta- 
tion a decided reversal is noted. The Atlantic border is depressed 
and much of the interior region seems not to have been subjected 
to further deposition from that time even to the present. 

d Mesozoic time. Again conglomerates, sandstones and 
shales were laid down upon an eroded floor. From their condition 
and lithology it is believed that they are partly of continental, flood 
plain, origin. The series is thick, generally assigned to the Triassic 
period and is extensively developed. During the time of accumula- 
tion and to some extent subsequent to it, there was extensive 
igneous activity pouring out and intruding basic basaltic matter in 
large amount. The Palisade diabase sill, and the Watchung Moun- 
tain basalt flows are the best examples. 

At a later time small faulting occurred making frequent dis- 
placements in this series. But mountain-folding has not again 
visited the region. Such breaks as there are, are of the nature of 
overlaps and disconformities rather than of the revolutionary his- 
tory indicated by a true unconformity. One of these intervals 
occurs in the Mesozoic between the Triassic and Cretaceous. Above 
it the thick series of Cretaceous shales, marls, sands and clays are 
developed. Succeeding this series a similar interval represents the 
earliest Cenozoic time. 

c Early Cenozoic time. The earliest Cenozoic (Eocene and 
Oligocene) has no sedimentary record within this region. 

There are small remnants of deposition representing Miocene and 
Pliocene time. Above these again the record is blank up to the 
time of the glacial invasion. 

/ Late Cenozoic time — glacial period. By some combina- 
tion of conditions not very well understood, the chief features of 
which no doubt are, — (i) continental elevation and (2) shifting 
of centers of precipitation and (3) modification in the composition 
of the atmosphere, a period of excessive ice accumulation was 
inaugurated. Ice finally covered immense continental areas and 



from its own weight by continuous accumulation spread out 
(flowed) from great central areas toward the margins. There is 
clear evidence of interruptions or advances and retreats of this 
general movement many times. But the same type of work and 
similar results were attained in each case. The chief features of 
this work was the moving of rock material frozen in the ice to long 
distances and the deposition of it again, more or less modified by 
its contact with the ice or by the effect of water upon its release, 
at other places and with entirely new associations. The tendency 
to ice accumulation was finally overcome to sufficient extent for the 
inauguration of the present condition of things. Whether it is a 
permanent change or only an interglacial interval is not clear. 
But the ice has withdrawn to the mountains and the polar north 
at the present time. It has not occupied the surface of this region 
probably within the last 40,000 years, and perhaps for a much 
longer time. 

5 Outline of geographic history — physiography 

The surface features of a country are the result of the working 
out of a long and complex series of processes with and upon the 
materials of the rock floor or bed rock. The relationship of surface 
features to the formations that occur in the rock floor and their 
stages of development, in short, an interpretation of their origin and 
meaning, constitutes geographic history or physiography. It differs 
little in essential character from geologic history, of which it is only 
a special branch, i. e. the history of surface configuration. And it 
can not be appreciated or understood except in the light of a 
thorough knowledge of stratigraphic and structural geology. In 
individual cases or particular regions the geologic knowledge must 
also be specific. 

a Early stages. Occasional glimpses of surface features, and 
some scattered facts about their development are to be gathered of 
older continental existence. Surface features characteristic of their 
time were developed in the great intervals between each successive 
period of continuous deposition. Traces of them are involved in the 
unconformities of the geologic column already shown in the discus- 
sion of geologic history. Hills, valleys, streams, shores and all the 
appropriate assortment of forms must have existed. But they 
could not have been like those of the present in many minor fea- 
tures — especially in arrangement and distribution — because the 
bed rock of those times had only in part reached the complexity of 



structure and composition now belonging to it. Many items of im- 
portance are indicated in some of these early periods. For ex- 
ample, the sea encroached on the land borders repeatedly from the 
westward — especially throughout Palezoic times, while in Meso- 
zoic and Cenozoic times the evidence of shif tings of sea margins 
is confined to the east and southeast borders, and likewise probably 
no near by place has been continuously beneath the sea. 

But the unraveling of these conditions is obscured by subsequent 
events. Land surfaces that once were, became covered by later 
sediments. The physiography of those times, Paleophysiography, as 
well as paleogeography, is therefore a difficult and intricate line of 
investigation. With these ancient* surfaces the dicussion of present 
features has little to do. Here and there the present surface cuts 
across and exposes the edges of an older one giving traces of the 
old profile ; but in most cases it is so distorted by the foldings and 
other displacements belonging to a later period that a restoration 
of the original continental features is a task fit for the most highly 
trained specialist. 

The surface as it now exists, and the rock floor modified only by 
the inequalities of the loose soil mantle, yields more readily to in- 
vestigations of origin and history. 

b History of present surface configuration. On some por- 
tions of the region there seems to have been no deposition since the 
close of Paleozoic time. Throughout most of Mesozoic and Ceno- 
zoic times, therefore, those regions probably have been continuously 
land areas (continental) and have been subjected to the agencies 
of erosion. This applies particularly to the Highlands region and 
the Catskills and the Shawangunk range and intervening country. 

What the surface configuration was like in the early stages is 
wholly unknown. In the beginning, mountain-folding — the Appa- 
lachian folding — was in progress and the features were probably 
those of partially dissected anticlinal folds. With the progress of 
erosion the Triassic deposits were accumulated along the eastern 
border, probably on the continental slopes. Subsequently, further 
elevation extended erosion over the Triassic areas also and the 
Cretaceous beds were laid down on the margin. The general lines 
of development have been the same from that time to the present. 
Each successive important formation less heavily developed and 
forming a band outside of and upon the older one — the whole now 
constituting a series of successive belts the oldest of which is far 
inland and the newest at the sea margin. 


6 7 

Therefore, when long periods of denudation are referred to, it is 
well to appreciate that this is especially applicable to the interior, 
that the sea margins are comparatively new, and that certain of the 
inland areas were suffering erosion long before the rock forma- 
tions that lie beneath and form the rock floor of the sea border 
districts were in existence. 

Cretaceous peneplain. It appears from studies of these problems 
in a broad way, and, drawing upon generalizations from continental 
features of a much larger field than that of the present study, that 
the continental region of which this forms a part must, in the 
earlier periods, have remained in comparatively stable equilibrium 
for an extraordinarily long time. So long a time elapsed that most 
of the area was reduced by erosion to a monotonous plain (pene- 
plain) at a very low altitude, probably not much above the sea 
(base level). Only here and there were there areas resistant enough 
or remote enough to withstand the denuding forces and stand out 
upon the general plain as remnants of mountain groups (Monad- 
nocks). Possibly the Catskill mountains of that day had such 

This reduction of surface feature it is believed was reached in 
late Cretaceous time. The continent stood much lower than now. 
Portions that are now mountain tops and the crests of ridges were 
then constituent parts of the rock door of the peneplain not much 
above sea level. This rock floor was probably thickly covered with 
alluvial deposits (Hood plain) not very different in character from 
the alluvial matter of portions of the lower Mississippi valley of 

Upon such a surface the principal rivers of that time flowed, 
sluggishly meandering over alluvial sands and taking their courses 
toward the sea (the Atlantic) in large part free from influence by 
the underlying rock structure. The ridges and valleys, the hills, 
mountains and gorges of the present were not in existence, except 
potentially in the hidden differences of hardness or rock structure. 
Such conditions prevailed over a very large region — certainly all 
of the eastern portion of the United States. This so called Creta- 
ceous peneplain is the starting point in development of the geo- 
graphic features of the present. 

Continental elevation. Following upon this period of stability 
and extensive denudation came one of continental elevation. How 
much above sea level this raise:! the areas under present discussion 
may not be determined, but that it was a sufficient amount to 



rejuvenate the streams and permit them to begin the sculpturing 
of the land in a new cycle of erosion is perfectly clear. As soon 
as the elevation and warping of the continental border made its 
influence felt in the increased activity and efficiency of the streams 
(rejuvenation) they began transporting the alluvium of their flood 
plains and to sink their courses through this loose material to bed 
rock. The final result of long continued denudation under these 
conditions in early Tertiary time was the removal of the loose 
mantle and the beginning of attack on bed rock (superimposed 
drainage). The streams formerly flowing on alluvium that had 
now cut down to rock found themselves superimposed upon a 
rock structure not at all consistent with their former courses. 
With the progress of erosion on this rock floor all these differ- 
ences of structure, such as the differences in hardness of beds, 
the trend of the folds, the strike of the faults, the igneous masses, 
etc., were discovered and the streams began to adjust their courses 
to them. Valleys were carved out where belts of softer rock 
occur, ridges were left as residuary remnants where belts of harder 
rock exist, and the surface (relief) took on some of the char- 
acter of present day lines. That is, the principal mountain ranges 
of that time were the same as those of today in position and 
trend ; but they had not so great apparent hight because the in- 
tervening valleys had not yet been cut so deep. The principal 
escarpments of that time were due to the same structural lines 
as those of today, only they have shifted somewhat along with 
the general retreat of all prominences by the forces of weathering 
and erosion. 

In the course of this work of sculpturing and the shifting of 
valleys and divides and escarpments and barriers into constantly 
greater and greater conformity with rock structure, it came about 
by and by that practically all of the smaller and tributary streams 
had so completely adjusted themselves to their geologic environ- 
ment that their valleys almost everywhere followed along the 
softer beds (subsequent streams), the divides were chiefly of 
harder beds, the trend of both were almost everywhere parallel to 
the strike of the rock folds and other structures (adjusted drainage) 
This undoubtedly involved in many cases a very radical change of 
stream course, and in some cases an ultimate reversal of drainage 
to such extent that tributaries were deflected inland against the 
course of the master streams and in some cases actually flowed 
many miles in this reversed direction before finding an accordant 
junction (retrograde streams). At least three of the streams of 



southeastern New York are still of this type — the Wallkill, the 
Rondout and the lower portion of the Esopus. 

But the larger rivers, the great master streams, of the super- 
imposed drainage system, in some cases were so efficient in the 
corrasion of their channels that the discovery of discordant struc- 
tures has not heen of sufficient inlluence to displace them, or re- 
verse them, or even to shift them very far from their original direct 
course to the sea. They cut directly across mountain ridges be- 
cause they flowed over the plain out of which these ridges have 
been carved and because their own erosive and transporting power 
have exceeded those of any of their tributaries or their neighbors. 
They are superimposed streams (not antecedent), they have, with 
their tributaries, settled down in the ancient plain, and, by their 
own erosive activity, have carved the valleys deeper and deeper, 
cutting the upland divides narrower and narrower until now only 
here and there a ridge or a mountain remnant stands with its crest 
or summit almost reaching up to the level of the ancient pene- 
plain on which the work began, if the transported matter could 
all be brought back and replaced in these valleys the old plain 
might be restored, but the work would immediately begin al! over 

Of these great master streams the Hudson is the only local rep- 
resentative [see Study of the Hudson River gorge in part 2]. 

Tertiary incomplete pencplanation. Such processes, if allowed to 
continue on a stable continental region, would ultimately reduce 
the land for a second time to a monotonous plain (complete cycle 
of erosion). The beginnings of such a plain would be made in 
the principal stream valleys upon reaching graded condition. Their 
lateral planation and the development of flat-bottomed valleys 
would begin at about the level that the plain would stand in the 
final completed stage. The difference of elevation between the 
ridge crests or hilltops and these flat valleys, i. e. between the old 
peneplain and the new unfinished one would be an approximate 
measure of the amount of the continental elevation that instituted 
the new cycle. 

But judging from such remnants of this later plain as are to 
be seen, the two, i. e. the old Cretaceous peneplain and the new 
Tertiary peneplain are not parallel. Toward the southeast, toward 
the sea, the older plain descends more rapidly than the younger 
and intersects it. Both pass beneath sea level in that direction. 
The difference between them therefore varies with locality from 



o feet to perhaps 2000 feet within the borders of the area (con- 
tinental tilting or warping). 

Late Tertiary rcclevation. Traces of such an intermediate and 
incomplete peneplain are to be seen in the compound nature of the 
large valleys of the present day. Most of them are essentially broad 
valleys into the bottoms of which narrower valleys and gorges are 
cut. The tops of the minor hills and ridges of the broad valleys 
represent the intermediate Tertiary peneplain that was interrupted 
in its development before completion (interrupted erosion cycle). 
The inner narrow valleys indicate that for the second time a re- 
gional elevation rejuvenated the streams and they began their 
work of cutting to a new grade. They have made a good begin- 
ning at this task, and as a consequence have carved some rebel 
in the old valley bottoms. These new streams have not yet reached 
a graded condition. 

When the glacial ice began to invade this region all of the surface 
features had had such a history. Leaving out of account minor 
fluctuations of elevation and depression, of which there may have 
been several of too transient character to make a lasting impres- 
sion on the topography, the stages become comparatively few and 
the general tendencies are easily understood. 

The measurable differences of elevation between the Cretaceous 
and Tertiary peneplains give some reasonable conception of the 
amount of the first continental or regional elevation. Concerning 
the altitude reached in subsequent regional elevation there is less 
certainty. None of the streams, not even the master streams such 
as the Hudson, reached grade, for it exhibits strictly a gorge type 
not only within the present land borders, but it is now known to 
show gorge development far beyond the present coast line. Judg- 
ing from the Hudson, therefore, it seems necessary to conclude 
that this continental region stood at a much greater elevation in 
some portions of the later period than had formerly prevailed. 
Probably the maximum elevation immediately preceded the glacial 

Conservative estimates as to the amount of elevation of that 
time in excess of the present would place it at not less than 2000 
feet. Much more than that is believed to be indicated, possibly 
5000 feet or more. 

In the meantime, the master stream, the Hudson and several 
of the tributaries cut into their valley bottoms to such extent as 
to make typical gorges so deep that their beds now, since the sub- 



sidence, lie much below sea level. The Hudson bed is of this 
character throughout its course from Albany to the Atlantic, and 
in the Highlands, 60 miles inland, the known rock bed at one point 
is more than 700 feet below sea level. 

In late glacial time there was still greater subsidence (50-100 
feet) than the present as is indicated by terraces above present 
water level and the deltas formed at the mouths of tributary 

Such in general outline is the history of successive conditions 
governing the topographic development of the rock floor. The suc- 
cession of periods of stability, elevation, stability again, reelevation 
and subsidence have had an effect on all sorts of formations, but 
the extent of the impress and its permanence varies greatly in 
the different districts. Tt is not possible to study these differences 
in detail here. They are the minor and special local characters that 
are in control at particular localities. In discussions of special 
problems some of these are taken up in more detail. But in each 
case the general history as outlined above, together with the modi- 
fying influence of known local structure and stratigraphic char- 
acter are the foundations of a working understanding [see Hudson 
River crossings, Moodna creek, Rondout valley, etc., pt 2]. 

Pleistocene glaciation. An additional modification and one largely 
independent of and largely inconsistent with the distribution of the 
smaller features of the rock floor is introduced by the glacial drift. 
It covers almost everything, but so unevenly as to largely destroy 
some of the detail. It is in places more than 350 feet thick (as 
in the Moodna and Rondout valleys') and in others it amounts to 
nothing. Tt covers the narrow ravines and gorges heaviest and 
has altered the courses of many of the smaller streams, the original 
channels being hopelessly buried. The result has been chiefly one 
of reducing the ruggedness of outline that prevailed along the 
newer gorges of late preglacial time. 

Besides this the usual surface forms characteristic of glacial de- 
posits, occur — -the kame, the drumlin, the esker, the hill and ket- 
tle topography of the terminal moraine, the overwash plain, the 
delta, the lake deposit and the gentle undulations of the ground 
moraine. These are superimposed on the rock floor features. Both 
are equally important to understand in the problems that have been 
encountered. Which set of factors is to be most regarded in a 
given case depends wholly upon the locality and the kind of en- 
terprise or work it is proposed to undertake. 



c Physiographic interpretation. Rock floor contour is an ex- 
pression of the differences in character and structure of the bed 
rock formations themselves, brought about by ordinary surface 
weathering and transporting agencies, varied in their action and 
effects only by certain differences in elevation above the sea. It 
is apparent therefore that it would be possible by careful observa- 
tion of surface features to gather data sufficiently definite to fur- 
nish a basis for suggestions about hidden and hitherto unknown 
or undiscovered structural and stratigraphic characters. But the 
application of it to practical engineering problems is a complicated 
and difficult matter. And this difficulty is nowise simplified by 
the occurrence of a drift soil that tends to obscure many of the 
more delicate features. For example, the later narrow stream 
gorges marking the stage of extreme regional elevation are com- 
pletely buried. Only an occasional stream like the Hudson has 
maintained its course unchanged and has begun excavating the 
channel again. But even in this case, as will be shown under a 
separate head, the work of reexcavation is only just begun and 
the amount yet to be done and the corresponding original depth 
of the gorge are wholly unknown. 

Certain surface features, however, are readable and, considered 
with due regard for all possible causal factors, give very useful 
suggestions. From them one obtains clews as to (i) the attitude 
nr relations of the hard and soft beds and the weak zones, (2) 
the dip and strike of strata, (3) the persistence of a formation, 
(4) the occurrence of faults. (5) the direction of the chief dis- 
turbances, (6) the resistance and durability of local rock types — 
in short the structural characters of all kinds because difference^ 
in the distribution of these characters have given the different topo- 
graphic forms and geographic areas. They have made the feature^ 
of the Highlands look different from those of the Catskills, and 
those of Wallkill valley different from the Croton. Because of 
the long train of conditions with which these surface features arc 
each involved and the structures that they indicate they become 
easily the chief factors in preliminary judgment of comparative 
practicability of rival locations, and are the most reliable guide to 
direction and character and extent of exploratory investigation for 
many engineering enterprises. 

d Physiographic zones. Tn summarizing the physiographic 
data it appears that the following belts or zones may be regarded 
as fairly distinct units : 

Plate 12 


Catskill and 
Oneonta sand- 
stone conglom- 

Sherburne flags 

Hamilton and 
Marcellus shales 

Onondaga lime- 

Esopus grit 

The Helderberg 


Hudson River 
shales, sand- 
stones and slates 

Wappinger lime- 


Storm King 


The Highlands 
"i gneisses 

|| The Catskill I 

Ashokan ^reservoir 

Hamilton escarp- 

Esopus creek 

High Falls 

Rondout creek 

Wallkill river 

Hudson river 

New Hamburg 
Wappinger creek 

Fish kill creek 
Newbn rgh 


Storm King 
Bull mountain 
Crows Nest 
Foundry brook 

Cold Spring 
West Point 

Relief map of the region from the Catskill mountains to the Highlands 
showing the principal physiographic features. (The original model shows 
also the areal and structural geology.) (Taken from model made in the 
physiographic laboratory of Columbia University by Messrs Billingsley, 
Gnmes and Baragwanath) 



(1) Coastal plain. A district underlain by Cretaceous and later 
rocks and confined to a part of Staten Island and Long Island, 
not exceeding 400 feet relief. This zone is characterized by den- 
dritic drainage, except a narrow belt on its inner margin which 
is a longitudinal valley of the " inner lowland " type. Long Island 
sound occupies the position of this old adjusted valley. 

(2) Piedmont belt. A district lying between the coastal plain 
and the Highlands. It is underlain chiefly by crystalline rocks and 
metamorphosed sediments. Not exceeding 800 feet relief. It is 
characterized by adjusted drainage obscured only by drift. The 
ridges and valleys trend northeast and southwest close together 
and with very little variation on the east side of the Hudson, 
while on the west side the gentle dips of the Triassic give broader 
and more unsymmetrical forms with dip slopes and escarpments 
wholly independent of the opposite side. The zone is essentially 
transitional between the simple forms of the coastal plain and the 
complex mountainous character of the Highlands. 

(3) Highlands. The rugged elevated zone formed by the crys- 
talline gneisses. Reaching elevations of 1600 feet. It is character- 
ized by irregular mountain masses and lofty ridges of a general 
northeast trend but with many prominent irregularities both of 
form and of drainage. The valleys are deep and narrow. There 
are many steep escarpments. It is a mountainous zone in which 
complex structures and rocks have led to the development of com- 
plex forms. The zone forms a sort of barrier 20 miles wide across 
the Hudson river which exhibits its most zigzag and narrow and 
gorgelike development in this district. 

(4) Appalachian folds. Characterized by folded Paleozoic rocks 
north of the Highlands. Reaching elevations of 1500 feet rarely 
— general relief 400-800 feet. North of the Highlands the relief 
is much less pronounced. The softer rocks of the early Paleozoic 
formations permitted the development of a broad valley with almost 
perfectly adjusted tributaries, most of which on the west side of 
the Hudson are reversed. The topographic forms give expression 
to the universal folding and faulting of the formations. It is 
essentially a transition from the complex mountain zone of the 
Highlands to the much simpler Catskill area. 

(5) Catskill Monadnock group. Characterized by undisturbed 
Paleozoic strata and very strong relief — reaching elevations of 
3500 feet. The eastern margin is an escarpment facing the Esopus 
and Rondout valleys which are adjusted to the gently dipping 
strata of that side. Over the rest of the district the beds lie so 



flat that drainage is essentially dendritic modified slightly hy joint- 
ing. The great relief of the Catskills is due wholly to erosion 
of flat but very resistant strata that withstood the destructive ero- 
sion of Cretaceous peneplanation and stand as residuary rem- 
nants even to the present time. The Catskills are therefore essen- 
tially a Monadnock group. In structure they are almost as simple 
as the higher portions of the cuesta of Long Island, and they hold 
the same relation to the forms developed by erosion out of the 
old Paleozoic coastal plain of the interior. 


Physiographically the most complex zone is midway in the region 
under discussion — i. e. The Highlands. This belt is bordered on 
both sides by less complicated zones of less relief, of more regular 
topographic forms and less obscure history — the Piedmont cone 
on the south and the Paleozoic folds on the north. The outer mar- 
gins are both simple, essentially eroded coastal plains with strata 
dipping away from the central belts and on which forms and drain- 
age lines characteristic of such history are developed. These outer 
zones are the coastal plain of Long Island on the south and the 
Catskill Monadnock group on the north. It matters little that they 
differ in age by almost half of the known geologic column. 




The group of studies assembled in this part are chiefly those that 
have required considerable exploratory investigation in connec- 
tion with the proposed Catskill aqueduct and that have furnished 
new data of a geologic character. In some cases the additional 
investigations have discovered new and wholly unknown structures 
or conditions and in all cases the features as now established are 
much more accurately known than would otherwise have been 

The benefits of the studies have been twofold and reciprocal. 
On the one side the practical planning of the enterprise has con- 
stantly required an interpretation of geologic conditions as a guide 
to locations and methods and on the other the extensive investi- 
gations carried on have given an opportunity for practical appli- 
cation of geologic principles under conditions seldom offered and 
the data secured in additional explorations serve to make the detail 
of some of these complex features now among the most fully 
known of their kind. Examples of such cases are (a) the series 
of buried preglacial gorges (as in the Esopus, and Rondout and 
Wallkill and Moodna valleys) and (b) the completed geologic 
cross sections (such as the Rondout valley, the Peekskill valley, 
Bryn Mawr, etc.) and (c) the numerous additions to the knowledge 
of local rock conditions (such as that at Foundry brook, Rondout 
creek, Coxing kill, Pagenstechers gorge, Sprout brook, and others). 

Almost every locality has its own specific problem and its own 
peculiar differences of treatment and interpretation of features. 
Nearly all of the studies here presented came to the attention of 
the writer and others 1 in the form of definite problems or questions 
involving an interpretation of geologic factors and an application to 
some engineering requirement. Some of these questions, as is 
pointed out more fully in part i, chapter 2, are (a) the location of 

1 Professor James F. Kemp of Columbia University and W . O. Crosby 
of the Massachusetts Institute of Technology and the writer constituted the 
regular staff of consulting geologists. 


7 6 


buried channels beneath the drift, (b) the character and depth of 
the drift, (c) the kind of bed rock, (d) the condition of bed rock 
for construction and permanence of tunnel, (c) the underground 
water circulation, (/) the occurrence of folds and faults, (g) the 
position of weak zones, (/;) the depth required for substantial con- 
ditions, and many other similar problems. 

These need not be treated in their original form. Indeed many 
of them have now ceased to be problems in any real sense, for sub- 
sequent provings have made them simple facts, and wholly new 
questions came to take their places. In some of the larger prob- 
lems, however, it is believed that a treatment which involves a dis- 
cussion of the original problem and the method of solving it, to- 
gether with the data thus secured and the final interpretation of 
geologic features as now understood or established will be more 
instructive than a mere enumeration of the collected results. 

So far as possible each problem is treated as a unit and fully 
enough to be understood by itself. But a general knowledge of 
local geology as outlined in part i is assumed. 



Surface topography constitutes the chief factor in determining 
the general course of the aqueduct. It is planned to control the 
water so that it will flow to New York city. There is therefore a 
gradual descent of aqueduct grade from 510 feet A. T. at Ashokan 
dam to 295 feet at Hill View reservoir. Wherever the surface of 
the country is approximately the same as the aqueduct grade for 
that district it permits of the so called " cut and cover " type of 
construction which is much cheaper than any other. Therefore, 
other things being equal, the position that will permit the greatest 
proportion of cut and cover work would have a decided advantage. 
So it is possible from any series of good topographic maps to lay 
out trial lines that are sure to be worthy of consideration. The 
topographic sheets of the United States Geological Survey and the 
maps of the New York Geological Survey are of great usefulness 
in such preliminary work. 

But a little field examination shows that there are many other 
features and conditions that materially modify even comparative 
cost and are still more important factors in consideration of per- 
manence and safety. Sometimes it is not apparent that a course 
has any objectionable features till considerable exploratory work 
has been done. Likewise a serious difficulty at one point may more 
than counterbalance advantages at some other, so that considerable 
portions of the line are finally shifted to a better average position. 
In the course of these preliminary explorations much valuable data 
have been secured that now relate to points a considerable distance 
off the present line. The information has, however, been necessary 
and useful. 

One of the cases of this kind where geologic conditions have 
had an almost controlling influence is involved in the choice of 
place of crossing of the Hudson river. It has involved a shift of 
the whole line between the reservoir and the Highlands. Diffi- 
culties encountered in finding a crossing of the Esopus also con- 
tributed to the argument favoring a shift of the line [see map of 
trial lines west of the Hudson] . One of the points where explora- 
tory work had reached definite results before the more southerly 
line was finally adopted is near West Hurley. Here wash borings 




were successfully put down through 
the fine sands and silts of the 
lower Esopus valley so as to give 
a fairly acceptable profile of the 
rock floor [see fig. 7] . Esopus creek 
in this portion of its course follows 
the Hamilton shales escarpment 
which forms a steep border on the 
west side, while the east border of 
the valley and floor are formed, 
by the underlying Onondaga lime- 
stone. Gentle westerly dips prevail 
for both formations, so that in the 
perfect adjustment reached before 
the glacial invasion a cross section 
would have shown a typical unsym- 
metrical valley — one side a gentle 
dip slope and the other a bluff de- 
veloped by the undercutting of the 
stream as it shifted against the 
edges of the shales. 

Results of exploration show that 
the valley is filled to a depth of 
more than 200 feet with silts and 
sands that are essentially overwash 
and glacial lake deposits. The flat 
surface further favors this explana- 
tion as had been pointed out before 
any explorations were made. Later 
observations in that portion of the 
Rondout valley which is a continu- 
ation of this structural feature indi- 
cate similar deposits as far soutn 
as the new line at Kripplebush, 10 
miles away. 

In this instance at West Hurley 
by careful measurement of dips on 
the Onondaga limestone and the 
Hamilton shales it was possible to 
estimate the approximate depth to 
which the Onondaga floor rock 
would pass by the time the base of 



the escarpment is reached. It was further helieved that the cov- 
ered portion is wholly drift-filled down to the Onondaga. It was 
easy therefore to estimate the approximate profile and suggest the 
point of greatest probable depth. The accompanying figure illus- 
trates the form and structure of this valley. Each valley has had 
in a smaller way a similar study and adjustment of location of line. 

The final result is shown on the accompanying map which indi- 
cates the course of the aqueduct as now being constructed. 



This is a special study of the Hudson river gorge 1 based upon 
explorations by borings at the several proposed crossings. Alto- 
gether 226 preliminary borings were made on 14 cross sections. 
The most important lines of borings are located at seven different 
points on the Hudson [see location map]. Four of them are in the 
vicinity of New Hamburg, lying not more than a couple of miles 
north and south of that village, while three others are located within 
the Highlands. [See comparative geologic study in following 
chapter.] The chief basis of information on all but one of these 
lines is the wash rig, a contrivance as already pointed out that gives 
rather incomplete data [see Relative Values of Data, pt 1]. On 
this account it is not possible to give the true bed rock profiles of 
the river canyon even approximately except at one location, i. e. 
the Storm King— Breakneck mountain line. An occasional diamond 
drill hole has been put down on some of the others and this has 
been done systematically at the Storm King location in a persistent 
effort to determine the gorge profile and bed rock condition. 

The work already done has proven that in the Hudson at least 
the wash rig borings give wholly unsatisfactory profiles. The holes 
do not penetrate the boulders and heavy glacial drift that is now 
known to fill the canyon. The profiles, however, that were drawn 
from this sort of data have some value. They indicate that bed 
rock is still lower and that the finer silts extend down to these 
depths. In some places there is a heavier filling of 400 to 500 feet 
below them before the rock floor is reached. 

Wherever the diamond drill has succeeded in reaching rock the 
formational identification has been made and the geological cross 
section is a little more complete. As a matter of fact, however, at 
almost every locality the structural relations are so complex or so 
obscure that they are still not fully known. The accompanying 
profiles and cross sections summarize the mass of accumulated data : 

1 Kemp, Prof. J. F. Buried Channels beneath the Hudson and its Tribu- 
taries. Am. Jour. Sci. Oct. 1908. 26:301-23. Some of the accompanying de- 
scriptions of river crossings follow closely this excellent summary of Hud- 
son river explorations from Professor Kemp. 



Fig. 9 Key map showing the locations of lines of wash borings 
forming the basis of the accompanying cross sections of- the Hud- 
son above the Highlands 


8 3 

i Points of exploration 1 

a Tuff crossing. This line is a half mile above Peggs point. 
Wappinger limestone forms the east bank of the river and Hudson 
river slates the western bank. There seems to be no abnormal 
structural relation of the formations. All data are from wash 
borings. The accompanying section gives the results. 

b Peggs point line. Peggs point is 2 miles north of New 
Hamburg. At this location Wappinger limestone forms the east 
bank and Hudson river slates the west bank of the river as in the 
previous case. The limestone dips gently westerly while the slates 
have a variable attitude. This is a normal relation and there is no 
direct evidence of any great structural break. A large number of 
wash borings have been made and five diamond drill holes were 
driven, three of them in the river. None indicate a greater depth 
than 223 feet, although there is a wide stretch, 1040 feet, not ex- 
plored by the diamond drill. This space must contain the deeper 
gorge if one exists here. From the known conditions at the 
entrance to the Highlands, 10 miles further down stream, where 
the channel is known to be more than 500 feet deeper, it may be 
rather confidently asserted that a deeper inner channel does exist at 
this point. 

c New Hamburg line. This line crosses the Hudson from 
Cedarcliff to the village of New Hamburg. The river is narrow — 
only 2300 feet. There are no drill borings within the river channel, 
but there is one on each bank. Both penetrate Wappinger lime- 
stone first and then pass into Hudson river slates beneath. How 
much of a gorge exists here is wholly unknown except in so far as 
may be judged from the wash boring. There are the same reasons 
for believing that a gorge exists as those noted for the Peggs point 

Structurally this line is probably the one of greatest complexity. 
It is however perfectly clear that the abnormal position of the 
slates and limestone on the east side of the river is caused by a 
thrust fault. A similar relation of the slates and limestone on the 
west side must be due to a like movement, but whether they are 
separated portions of the same structural unit or of two adjacent 
ones is not clear, although they are probably distinct 

1 All of these explorations on the Hudson river have been under the direct 
supervision of Air William E. Swift, division engineer, in charge of the 
Hudson River division. 

8 4 



Tuff ffff/vGS 5cc/,<,nUK X 

Fig io Cross sections of the Hudson river north of New Hamburg and of Wappinger creek 
based upon wash borings. [For locations see key map, fig. 9] 



Five lines of wash borings were followed, and the results of these 
are indicated in the accompanying figures. A maximum depth of 
263.5 f eet is shown by these wash borings. 

d Danskammer line. This line is about a mile south of Xcw 
Hamburg. Two lines of wash borings were made, reaching a 
maximum depth of 268.5 f eet - I" this case slates standing almost 
vertical form the east bank and limestone dipping gently eastward 
the west bank of the river. Whether there is a deeper gorge or a 
more complex structure here is wholly unknown. 

Of the three remaining lines, all of which are within the High- 
lands, that one projected between Storm King mountain on the 
west and Breakneck ridge on the east has been much the most 
thoroughly explored. It is known as the Storm King line. The 
other two have seemed to merit less attention. One crosses the 
river from Crows Nest mountain to Little Stony point and Bull 
mountain just north of Cold Spring, and is known as the Little 
Stony point line. The other crosses at Arden point about a mile 
south of West Point and Garrison. 

e Arden point line. Only wash borings were made. A 
maximum depth indicated by this method is 220 feet. Structurally 
this location appeared to have disadvantages, and although the 
evidence as to bed rock conditions is confined to the natural out- 
crops, there is no doubt but that it has objectionable features of this 

The Hudson follows closely the structural control in this portion 
of its course. These structural elements include the foliation, the 
bedding of the original sediments, the subsequent shearing zones, 
and the strike of folds and faults. Crushed and sheared zones are 
nowhere in the Highlands seen so extensively developed as on the 
islands and the east bank of the Hudson in this, the central portion 
of its Highlands course. The river is very narrow, being only 2120 
feet on this line. 

f Little Stony point line. The river here is 2360 feet wide. 
The rocks on each side are similar and give no clue to possible 
depths of channel. Less than 200 feet was reached by the lines of 
wash borings. Three drill borings penetrated the stony or bouldery 
river filling somewhat deeper — one near the center reaching 322 
feet. None, however, reached bed rock. 

g Storm King crossing. Extensive exploratory work has 
been carried on at this point, both on the banks and in the river. 
Wash borings as usual have given poor results. Two diamond drill 
holes were run at an angle toward and beneath the margins of the 




i P£Ges Po/a/t 5o</TH /?f?AI6£ Section O.P J 

Fig. II Cross sections of the Hudson river near New Hamburg based on wash borings 
[For locations see key map, fig 9] 



river, and in addition a working shaft suitable for permanent use 
has been started on each side of the river. These have thoroughly 
explored the rock character to a depth of about 800 feet, it has 
proven to be of constant type, a gneissoid granite, affected by 
moderate amount of jointing, shear movements and occasional dike 
intrusion. The two sides are alike, the rock in depth is com- 
paratively free from water, nearly all coming from the adjacent 
surface drainage. 

Persistent efforts have been made to use the drill in the river to 
explore the rock channel, but with meager results. The difficulties 
to be overcome in drilling in this tidal river to the necessary depth 
are probably greater than have even been encountered in any 
similar undertaking. The disturbance presented by the current, the 
tide, the depth of water, the drift filling above the rock channel, 
and the traffic in the river are a constant menace. The complex 
character of drift filling in this gorge, especially the occasional 
heavy bouldery structure, makes it necessary to reduce the size and 
recase the holes repeatedly. But in this regard the work has 
suffered less actual loss than by the menace of river traffic. 
Several times after the greatest efforts had been put forth in 
pushing the drills deep into the gorge a helpless or unmanageable or 
carelessly guided steamer or scow has wrecked the work. In this 
way some of the most critical locations have been lost together with 
many months of labor. 

The results are shown on the accompanying drawings. 

It is worth noting that of those holes located far out in the river 
channel only two have reached bed rock. Even these two have 
penetrated the rock so little distance that there might be still some 
doubt of permanent bed rock. The fact, however, that the rock 
found is of the right type, i. e. like the walls of the gorge, leads to 
the conclusion that the bottom was actually penetrated. Neither of 
these holes are in the middle of the river, and, although the 
maximum depth of 608 feet was reached by one of them, the central 
portion of the buried channel proves to be still deeper. One hole 
located near the middle was able to penetrate to a depth of 626 feet 
without striking bed rock. But it was finally lost. The latest 
results are from a boring that has reached a total depth 1 (January 
1, 1910) of 703 feet, the last 8 feet of which was believed by the 
drillers may be in bed rock. All above is drift and silt. 

1 Subsequent exploration has proven that the bottom of the old channel 
lies still deeper. This boring has been pushed to a depth of 751 feet with- 
out yet touching bed rock (Oct. 8, 1910). 



J r ig. 12 Cross sections of the Hudson river at four points between Danskammer Light and 
New Hamburg [see key map, fig. 9, for locations] 



8 9 

2 Discussion 

The present facts therefore indicate that the buried Hudson 
channel is more than 700 feet deep between Storm King and 
Breakneck ridge. Furthermore this is more than twice as great 
depth as has been found (so far as yet tested) at any other point 
either above or below this place. Although data of this kind are 
scarce yet there are two other borings that have given surprising 
results — -(a) at Peggs point and (/?) the Pennsylvania borings at 
New York city. 

Peggs point. At this place, where studies were made for a 
possible crossing, a hole 700 feet from shore struck rock at 223 
feet and the unknown space or interval within which it is possible 
for a channel to lie is less than 1040 feet wide. This is about 10 
miles above the Storm King crossing and in much softer rock 
(Hudson River slates). Yet the Storm King gorge in granite is 
deeper than that (deeper than 223 feet) for a width of nearly 2500 
feet. Of course, there may be, and probably there is, a much 
deeper channel at Peggs point within the 1040 feet unexplored 
space. But even so there is a remarkable discrepancy in width of 
gorge at these two points that must be accounted for in some other 
way than simple stream erosion. 

The Pennsylvania borings opposite 33d st., New York city. 
The data gathered by the Engineers of the Pennsylvania Tunnel 
Company in their explorations for tunnel from 33d street, Man- 
hattan, to Jersey City, have recently been made public. There are 
six holes into rock. Their positions and depth to rock bottom are 
given below : 

a 800' from New York bulkhead 190' to bed rock = aplite 
b 1000' from New York bulkhead 290' to bed rock = hornblende 

c 2180' from New York bulkhead 300' to bed rock = chloritic 
and serpentinous rock. 

d 2350' from New York bulkhead 26o'( ?) to probable boulder = 
jasper breccia 

c 3300' from New York bulkhead 270' to bed rock = arkose 

/ 13700' from New York bulkhead 225' to rock=brown sand- 

There are other shallower borings on both sides of the river. 
Those on the Manhattan side are represented by several different 
facies of Manhattan mica schist and granite and pegmatite in- 



trusives, while the New Jersey side is represented by different 
varieties of arkose and gray and brown sandstone belonging to the 
Newark series. 

It should be noted that although only one hole marks rock bottom 
as low as 300' (that one situated 2180' from the New York bulk- 
head about the middle of the river), yet there is at least a 1100 foot 
space on each side which is essentially unexplored, and within one 
of these spaces there may be a deeper gorge. 

The cores taken from the east side of this middle zone belong to 
facies of the Manhattan schist formation, while those on the west 
side .belong to the Newark series. The middle one, however, is 
essentially a soapstone or serpentine and may be a continuation of 
the Hoboken serpentine belt. In any case, it belongs in age to the 
older series of formations. 

It is certain that here again, 50 miles below Storm King locality, 
a very deep gorge, if one exists, must be comparatively narrow. 

Submarine channel. It is worth noting in this same connec- 
tion that a submerged gorge has been mapped by the Coast and 
Geodetic Survey on the continental shelf from the vicinity of 
Sandy Hook to the deep sea margin, a distance of more than a 
hundred miles. This is interpreted by Spencer 1 and others with 
apparently sound argument as the lower portion of the old pre- 
glacial Hudson gorge formed during an epoch of great continental 
elevation. The outer portion of this submerged gorge is very deep. 
That section near shore is shallow and obscure. It has been 
assumed that this obscurity and shallowness is due to offshore and 
river deposition, filling the channel with silt. No better explanation 
is yet forthcoming. But even here the width of the submerged 
gorge is suggestive. In very much softer sediments than any en- 
countered in its whole course on present land, and in a part of its 
course from 50 to 100 miles below the other sections, the river has 
cut a gorge only 4000 feet wide at top and 2000 feet deep within 
a broader valley 5 miles wide. In its deepest known part the 
proportions are 10,000 feet in width at top to 3800 feet in depth. 

From this it would appear that the inner gorge type of develop- 
ment is characteristic of the Hudson, and that it was originally an 
exceedingly narrow one compared to the present river width, indi- 
cating rapid erosion during a brief and comparatively recent epoch. 
This submerged continental margin condition is favorable to the 

1 Spencer, J. W. The Submerged Great Canyon of the Hudson River 
Am. Jour. Sci. 1905, v. 19. 



assumption that there are narrower, still deeper channels within the 
unexplored spaces both at New York city and at Peggs point. 

The only known exception and the one really surprising section 
is the Storm King crossing. It is too wide, considering the profiles 
at Peggs point and at New York city for simple normal stream 
erosion. That is clear enough. But a still more difficult question is 
whether it is also too deep. It is much deeper than any known 
section above or below for a distance of 50 miles. 

There appears to be only one satisfactory explanation of this 
abnormal width of the deeper section and that is by glacial erosion. 
Just above Storm King is the wide bay opposite Cornwall and 
Newburgh. The few glacial scratches observed trend about s. 15 e. 
The ice therefore moved to the east of south, and it is noted that the 
course of the river is about the same. The northern front of Storm 
King mountain is steep and trends east and west while the northern 
front of Breakneck mountain trends southwest. It would appear 
therefore .that these slightly converging mountain fronts served as 
sort of a funnel into which the ice was forced from the wide gather- 
ing ground immediately above, and through which there may have 
been a tongue or stream of ice of more than average power and 
efficiency moving almost in direct line of the present course of the 
river. It is reasonable to expect that these conditions would favor 
more than average glacial erosion. 

3 Storm King-Breakneck mountain profile 

It is practically impossible to draw a complete profile for the 
Hudson river gorge at any point in its lower course. Even at Storm 
King mountain or New York city or at Peggs point, at each of 
which places considerable exploratory work has been done, only the 
broadest features are known. Nevertheless, several things have 
been proven and they are worth considering in this question. They 
may be summarized as follows : 

a. If there is a very deep gorge at Peggs point (deeper than 250 
feet) it can not be over 1000 feet wide. 

b If there is a very deep gorge at New York city (deeper than 
300 feet) it can not be over 1200 feet wide. 

c At Storm King, located between the other two and in harder 
rock than either of them, a gorge at least 400 feet deep is proven 
to have a width of more than 1500 feet. 

It is certain that simple stream erosion could not account for 
such a difference of cross section. There is no doubt but that en- 




larging by ice so far as widening is concerned is practically proven. 
If may also be overdeepened, by which is meant that it may have 
been gouged out deeper than could have been done by a stream of 
water alone. 

If ice action then be granted, the profile ought to be and prob- 
ably is essentially an ice valley profile, i. e. of a more or less U- 
shape, rather than of typical stream erosion form. It is certain 
also in this case, if glacial overdeepening is admitted, that there can 
be no stream notch in the bottom of it. The significance of this 
lies in the probability that the floor is approximately the same level 
on a considerable portion of the bottom, so that when once the 
margin of this floor is touched the gorge as a whole is thereby 
determined for depth. 

After plotting the borings data and relying upon the factors that 
seem to be most firmly established, it appears that the following 
statements are as definite as the facts will warrant : 

a The average slope of the Storm King side of the valley above 
river level is nearly 38 , and this is in several steps or sections of 
steeper and flatter slopes. The Breakneck side is about the same. 

b The average slope of the Breakneck side of the gorge below- 
present water level (the side on which alone there are enough data 
to plot a fairly good curve) does not vary much from this same 
value [see accompanying profile] . And it is also in steeper and 
gentler slopes, apparently a series of U-shaped forms set one inside 
the other, each inner one deeper than the next outer one. Each suc- 
cessive inner step is approximately 300 feet deeper than the last 
and 1000 feet narrower. 

It is certain that this sort of profile is not as simple as at first 
appears. The surprising feature is the close approximation of the 
slopes above and below present river level. In view of the fact 
that glacial widening has been practically proven, as shown before, 
not much importance can be attached to this uniformity or simi- 
larity of slope. Ordinarily such a persistence of slope would be 
taken to indicate simple stream origin, but having abandoned that 
hypothesis, the value of the angle as a factor in estimating prob- 
able total depth is lost. In short, one can not assume that the 
deepest point is indicated by the intersection of the slopes of the 
two sides. 

But there is one feature that is at least suggestive. That is the 
uniformity of the succession of steps and slopes. It was noted 
above that each successive inner one is about 300 feet deeper and 
1000 feet narrower. If this uniformity and proportion is main- 



tained for the next inner one — inside of holes no. 10 and no. 22 
— there would be room for only one more and its approximate 
depth would lie somewhere between 800 feet and 900 feet below 

Recent drilling has shown a marked difference between holes 
no. 10 and no. 22. Hole no. 10 located 500 feet southeast of no. 
22 is nearly 100 feet deeper. Since no. 10 is nearly straight down 
stream this discrepancy is disturbing. But if one considers the 
distance of each from the east bank it is noted that no. 10 is 900 
feet out and no. 22 is 800 feet. Hole no. 10 is thus about 100 
feet nearer the middle of the stream and allowing for this addi- 
tional distance according to the profile as known it ought to be 
at least 70 feet deeper than no. 22. This corrected difference then 
of 30 feet does not seem to be of much importance. 

Summary. Everywhere in its lower course the Hudson ex- 
hibits the character of a narrow gorge, sometimes of a gorge within 
a gorge, most of which is either submerged or buried several hun- 
dred feet. 

Depths of 200 to 300 feet are average and for the last 60 miles 
of its course represent widths of 1000 to 3000 feet. 

Greater depths are believed to be maintained continuously within 
a narrower inner notch, but of this there is no conclusive proof 
and very little evidence outside of a few Storm King borings. 

The Storm King-Breakneck notch is over 751 feet deep. But 
it is abnormal at least in width and probably also in depth, due to 
ice erosion. 

The conditions indicate ( a ) rapid stream erosion while the con- 
tinent stood much higher than now, (b) glaciation which enlarged 
the gorge in at least a few places and filled it with rock debris 
and later with mud during submergence, (c) finally an emergence 
with minor oscillations and erosion to the present time. 

4 Origin of the present course of the Hudson 

The course of the Hudson is in most respects no more abnormal 
than that of the Susquehanna. Both flow across mountain ridges 
in such manner as to indicate their superimposed character. Both 
date back to the Cretaceous peneplain. But the striking feature 
of the Hudson is its straight course. As Hobbs and others have 
pointed out, the river is abnormally straight for more than 200 
miles — and this in spite of the fact that it crosses the bedding and 
other structures of the country rock at nearly all points at an 

9 6 


oblique angle. Such conditions are especially notable south of the 
Highlands where the Hudson cuts at a low angle across the ends 
of a succession of complex folds of the crystalline metamorphics 
for 30 miles to New York city. But this is true only of the east 
side of the river. The west bank is an almost unbroken uniform 
escarpment of the Palisade diabase intruded sheet underlain by 
Newark sandstones, which if laid down upon a pretty well planed 
Pretriassic surface might easily control the Hudson, and which 
would not differ from its present course. 

The most evident exception to this is the course of the river 
from Ploboken to Staten Island. Instead of following the line of 
contact between the crystallines and Triassic formations, the river 
cuts through the crystallines leaving large masses of serpentine 
and associated schist on the west side. This together with the 
behavior of the river in cutting across the strike farther north 
near the Highlands is believed to strongly favor the fault theory 
of location especially south of the Highlands. The same condi- 
tions would be favorable to the development of a narrow gorge 
and perhaps a very deep one rapidly eroded along the crush zone 
of the fault. 

From the northern entrance to the Highlands to Haverstraw bay, 
where the Palisades arc reached, the stream course is not by any 
means straight, but shifts from longitudinal structure to cross 
structure alternately in a zigzag manner. North of the Highlands 
the course is more direct again. On the whole the present explora- 
tions have added little to the facts bearing upon this question. 
Faults crossing the river arc common and easily recognized. Oc- 
casionally one appears to pass into the river gorge at a very small 
angle and not reappear. In a few places, especially in the High- 
lands, the course does not seem to be consistent with the hypothe- 
sis of a large fault line. It is to be expected that further work at 
the Hudson river crossing will add materially to ithe facts relating to 
the structures within the gorge. 



General statement 

This is essentially a study of the geologic features and condi- 
tions shown by exploration to have an important influence upon 
the choice of river crossing for the aqueduct. In the beginning it 
was possible to consider that any point between Poughkeepsie and 
New York might furnish a crossing. The early preliminary in- 
vestigations showed that it would be desirable to cross either above 
or within the Highlands and subsequent exploratory work throws 
light on different possible locations in these regions. Fourteen dif- 
erent lines were tested by wash borings. Later some of these were 
tested by diamond drill. As data accumulated it was possible to 
eliminate many of the trial lines and the more detailed and critical 
studies became confined to a few important possible crossings. 

In making a comparison of them as to geological environment it 
is evident that they fall into two distinct groups 1 [see fig. 15]- 
One, that may be designated the " New Hamburg " group is rep- 
resented by the " Peggs point, ' " New Hamburg," and " Dan- 
skammer " lines and is characterized by a series of much folded, 
faulted and crushed sedimentary rocks, chiefly slates, limestones 
and quartzites. The other, that may be called the Highlands group, 
is represented by the " Storm King," " Little Stony point," and the 
"Arden point " lines and is characterized by crystalline metamor- 
phic and igneous rock of a much older series. 

A judgment as to the most desirable crossing involves the selec- 
tion of one of these groups chiefly upon general geologic features, 
and finally a selection of a particular line upon minor differences of 
materials or structure. 

In the first place it seems necessary to consider, for each group, 

1 There have been other suggestions for crossing the Hudson river, 
farther upstream and farther down than these — one being at New York 
city — but none have had sufficient claim to attention to encourage much 
detailed work or so careful consideration as those here discussed. 

A shift of position of the Hudson river crossing involved a correspond- 
ing shift of a large section of the northern aqueduct line. The first choice 
of location occasioned a shift southward of all that portion between 
Ashokan reservoir and the Hudson. 




the whole length of pressure tunnels whose position would be modi- 
fied by a shifting of river crossing. This is because the aqueduct 
will approach the Hudson with nearly 400 feet head — i. e. 400 
feet above river level or with an equivalent pressure. For this 
reason it is considered necessary to plan a rock pressure tunnel 
beneath the river which can deliver the water at nearly the same 
elevation again on the east side. 

Thus any one of the " New Hamburg group " involves a contin- 
uous pressure tunnnel reaching from the margin of Marlboro moun- 
tain to Fishkill range, a distance of approximately seven miles, 
while any of the " Highlands group " permits the substitution of 
two more or less separate siphon tunnels (Moodna creek and Hud- 
son river) of considerably less combined total length. 

A reliable conclusion as to the choice of crossing is probably best 
reached through a comprehensive understanding of the geologic de- 
velopment of the region together with a consideration of specific 
local conditions. With this end in view a condensed outline of 
geologic history, so far as it bears upon the questions at issue, is 
inserted. But for a more comprehensive discussion of these matters 
the reader is referred to the explanatory chapter of part 1. 


This particular locality, including as it does the Highlands of 
the Hudson and the district lying along its northern border, is one 
of the most complicated stratigraphically and structurally to be 
found in the entire region. The strata represented include more 
than half the total geologic scale reaching from the oldest sedi- 
ments following the Archean up to and including a part of the 
Devonic series [see pt 1]. The rock types include granites, diorites, 
gneisses, schists, marbles, serpentines, slates, quartzites, sandstones, 
limestones, shales, and, less extensively, other varieties. And the 
region bears the evidence of no less than three periods of mountain- 
making disturbances, each in its turn adding to the succession of 
foldings, faultings and unconformities. 

The oldest formation is a crystalline gneiss — a characteristic 
rock of the Highlands. It represents an ancient sediment that has 
been completely recrystallized during some of the earlier mountain- 
making period. It is older than the Cambric. Interbedded with 
it to a limited extent are quartzite beds, ancient limestones (now 
usually serpentinous in character) and schistose beds ; and in it are 
many igneous injections, mostly granites of various types. All 




these igneous injections are therefore younger than the gneiss and 
are very large and abundant in certain cases. The granite of Storm 
King, Crows Nest and Breakneck ridge belongs to this type. 

Following the sedimentary cycle represented by the above series, 
and perhaps others not now preserved, the region was folded into 
a mountain range, the series was extensively metamorphosed and 
passed through a long period of erosion during which it was again 
reduced to sea level position and began to accumulate a new series 
of sediments. 

The lowest beds occurring upon this foundation are sandstones, 
now changed into quartzite. In places they are conglomeritic, and 
may now be seen projecting into the valleys along the Highland 
border. This formation is of Cambric age, and is from 200 to 600 
feet thick in favored places. It forms an almost continuous belt 
along the north side of the Highlands except where cut out by 
faulting, and extends with similar breaks beneath the later sedi- 
ments northward. This quartzite is known as the " Poughquag." 

Upon the quartzite of this series there was developed a succes- 
sion of limestone beds at least 900 to 1000 feet in thickness. This 
formation is known as the " Wappinger " and includes some beds 
that are of Cambric but for the most part of Ordovicic age. 

The final member of this series is a shale and shaly sandstone 
in places changed to slate. It is quite variable in actual character 
and has a great thickness, never yet successfully estimated, but 
probably several thousand feet. This is the so called " Hudson 
River slate " series. In this region they are of Ordovicic age. 

This is the succession which the proposed Hudson river lines 
has to penetrate in a pressure tunnel. Later Siluric and Devonic 
strata lie in the immediate vicinity of this alternative line, but add 
no complication to the problem as it now stands. Therefore no 
other formations need be considered except the glacial drift. This 
covers almost every rock surface and is deeply accumulated in 
some places, notably in the narrow gorges and valleys, obscuring 
the finer original topographic lines. 

A summary of the history of the formations chiefly involved in 
this problem with a suggestion of later erosion activities may be 
tabulated as follows : 

r Glaciation 
Erosion (interrupted) 
Elevation (rejuvenation) 






' Erosion to peneplain 


A long interval including two mountain-making epochs 
and at least one period of general sedimentation 


Ordovicic ^ ^ u< ^ son Ri ver slates 
| Wappinger limestone 
Cambric j Poughquag quartzite 


A long interval including mountain folding, igneous 
injection, erosion, and perhaps other sedimentations 

' The metamorphosed schists, limestones, quartzites 
etc., together with accompanying intruded igneous 
masses — forming the basal gneisses of the High- 

The evidence of such succession and history gathered from the 
scattered outcrops of rock in the immediate area, is nowhere better 
shown than in the field covered by this investigation. 


When such outcrops as are known are plotted and organized, sev- 
eral important facts become clear. 

1 The folds run with remarkable persistence northeast and south- 

2 The succession in many places is not normal. Often a whole 
formation or even two of them are missing and formations that 
should be separated are brought side by side. Faulting therefore is 
prevalent and the occurrences show that these large fault lines 
usually run northeast and southwest. 

3 A consideration of the dips of the strata shows that most of the 
folds are overturned as if pushed by some general movement from 
the southeast. 

4 This same movement causes the faulting to be largely of the 
overthrust type, and in some cases the lateral displacement attained 
in this way may possibly be several thousand feet. 

5 Isolated " islands " of the older rock formation appear out in 
the later sedimentary area. They all seem to belong to prolonga- 
tion of the ranges of the Highlands and their abundance undoubt- 



edly complicates the underground structure throughout a consider- 
able belt. / 

6 The Highlands area terminates in a serrate margin which, in 
the latest thrust movements from the southeast, must have created 
very unequal distribution of stresses within the slate-limestone 
region to the north causing additional cross folding and faulting. 
For the most part these can be traced only a short distance before 
losing their identity. 

In a mountain folding movement, the uppermost rocks are most 
broken and displaced or crushed while those of greater depth may 
be bent or uniformly folded or even recrystallized. It would ap- 
pear that this latter was the condition of the Highlands rock series 
during its earlier history. And even in the latest movements its 
lines appear to be less radically disturbed than the slates and lime- 
stones to the north. Most of the disturbances that invite serious 
consideration belong to the latest period of these mountain-making 

Comparison of routes 
i New Hamburg group. This group of crossings is in the 
later sedimentary series. Hudson River slates and Wappinger 
limestone are the chief formations. But within the southern third 
of the tunnel, at least, the underlying Cambric quartzite and the 
older Highland gneiss would be cut — the quartzite possibly three 
times. The succession therefore will be of considerable complexity 
as a whole. 

All of the formations involved are thrown into very steep dips 
at most places and are consequently liable to rapid and unexpected 
changes — some of which probably do not show at the surface. 

There are several fault lines belonging to the major northeast 
and southwest series to be crossed by such a tunnel — one of them 
in each case being met at considerable depth and beneath or adja- 
cent to the river. These faults besides being the weakest zones of 
rock as a rule, are in addition the most unstable in any possible 
future earth movements. Although there is no evidence of recent 
displacement along these lines, still such a thing is always possible 
and recent serious effects of this kind on the Pacific coast suggest 
caution. It is manifestly advisable, if possible, from every stand- 
point to avoid crossing several of them. 

In the field there are numerous springs of very large flow along 
many of the limestone borders. The concentration of them to 
these situations in addition to the occurrence of an occasional sink- 



hole, leads to the conclusion that they are more intimately depend- 
ent upon the limestone structure for their existence than upon the 
glacial drift or any superficial factor. Their abundant flow, some- 
times on high ground, indicates rather extensive structural con- 
nections and this is believed to be the limestone bed itself and that 
such flows would be encountered also in depth. The occurrence of 
sinkholes suggest also possible solution channels and cavities and 
distant outlets. The types of rock to be encountered on the lines 
represented by this group are easily workable. Among them all 
the Hudson River slates is probably the most satisfactory from any 
standpoint. It is generally easy to penetrate and has a capacity for 
healing its own fractures. For this reason it can be considered good 
ground, tight and safe. But a considerable distance of the tunnel 
can not be kept in slate — perhaps even more of it than can be 
proven from surface observations. The other formations are con- 
siderably less satisfactory. The limestones are in places shattered 
and are liable to abundant flow of water. The quartzite is ex- 
tremely hard, as difficult to penetrate as granite, and where crossed 
by the faults is probably not healed at all, while the gneiss is doubt- 
less of similar character to that of the Highlands crossings to be 
discussed later. 

Only minor modifications result from a choice of the individual 
crossing, whether " Peggs point," " New Hamburg," or " Danskam- 
mer." In one of them, New Hamburg, it would appear possible to 
cross the actual river section wholly in slates. This seems to be the 
reasonable conclusion from the diamond drill boring at Cedar Cliff. 
But even that line necessitates crossing at least two fault contact 
lines immediately at the east bank and beneath Wappinger creek at 
depths not immensely less than that below the river itself and both 
wholly within the range of influence of the river waters. It would 
appear therefore that the situation is not materially altered in the 
present discussion, no matter which particular crossing of this 
group is considered. 

2 The Highlands group [see cross section]. In this group of 
crossings there are two separate features to consider, (a) the 
Moodna creek valley which these lines all cross, and (b) the Hud- 
son river itself. Their characteristics are as follows : 

a Moodna creek [sec separate Moodna creek discussion]. So 
far as known Moodna creek can be crossed almost wholly in slate. 
It is possible that the underlying limestone may come near enough 
to the rock floor of the valley to be penetrated but there is little 



direct evidence of it. The ancient valley is deep and probably 
marks a line of displacements which can not be avoided, no matter 
what route is chosen. The fault contact at the border of the High- 
lands is not expected to prove troublesome as it seems very tight 
at the exposures seen. The buried granite ridge (a continuation of 
Snake hill) which underlies the western end is now known to come 
within the limits of the tunnel and adds one more complication. 

Except for the fact that the ancient Moodna valley is deep and 
filled with heavy drift that is unusually difficult to prospect, there 
would seem to be no source of special trouble. It has no lines of 
weakness that are not also present in the more northerly districts 
and the tunnel has chances of crossing them under more advan- 
tageous conditions without so much complication with the lime- 
stone series as characterizes the New Hamburg group. 

b Hudson river. Among the Highlands group of crossings there 
is considerable difference of structure dependent upon the exact 
location of the crossing. The conditions that prevail may be sum- 
marized as follows : 

(1) Storm King location. This is wholly in massive and gneissoid 
granite. The rock is the most massive and substantial body of 
uniform type found in the Highlands. The course of the river 
indicates some weakness in that direction. This weakness may be 
some minor crushed zone or even the jointing alone that prevails 
throughout the exposed cliffs. But there is no direct evidence of 
faulting, cutting the line and such crushing as may be encountered is 
believed to have originated at such depth and under such conditions 
as to cause no large disturbance. The freedom of this formation 
from all bedding structures and natural courses of underground 
water circulation on a large scale is an additional factor. There is 
absolutely no other place, within the region, where the Hudson river 
can be crossed from grade to grade in good ground of a single type 
with so great probability of avoiding all large lines of displacement. 

(2) Little Stony point location. The conditions that prevail at 
this point are similar to those that characterize the Storm King line. 
The only known difference is in the considerably more shattered 
condition of the granite, especially on the west shore at Crows Nest. 
It is estimated that this crossing is less favorable by reason of just 
this poorer condition of the rock and the somewhat greater yielding 
to regional disturbances that it seems to indicate. 

(3) Arden point or West Point location. On this line the river 
would be crossed in the gneiss series proper instead of in granite. 


It is largely an ancient stratified series much metamorphosed con- 
taining belts of interbedded limestones, quartzites, and schists, in 
addition to the more substantial feldspathic gneiss. The eastern 
bank of the river bears also abundant evidence of extensive crush- 
ing and shearing and is believed to indicate a displacement in this 
zone. For these reasons the West Point crossing is considered an 
unfavorable route compared to either of the others of the High- 
lands group. 

Summary. In a comparison of the geologic features that are 
of most importance in contrasting the possible routes for the Hud- 
son river crossing the following points are considered of most 

1 The New Hamburg group of crossings involves (a) the long- 
est tunnel, (b) the more complicated structures, (c) the greatest 
number of known faults, crush zones, and related disturbances, 
(d) the more variable series of rock types to be penetrated, (e) 
the greater tendency to encounter heavy underground water circula- 
tion, (/) the greater probable susceptibility to disturbance from 
future earth movements, and (g) the greater number of uncer- 
tainties of rock relations. 

2 In contrast the Highlands group admits of (a) shorter total 
tunnel length, (b) the most profound fault lines of the district are 
crossed either in high ground or are avoided or, because of the 
rocks involved, promise the least possible trouble, (c) the Hudson 
river itself can be crossed in a single formation with probability of 
avoiding lines of largest structural weakness confining the greatest 
pressures and deepest tunnel work within the most uniform and 
substantial rock of the whole region. 

There are, of course, many unknown or only partially known 
features obscured beneath the covering of drift or lying beneath 
the river itself ; but, however many there may be, it is not believed 
that they can materially change the general situation. The major 
characteristics are so well marked that any addition to those already 
known would in all probability increase the difficulties of the New 
Hamburg group of routes at least as much as and perhaps more 
than those of the Highlands group. 

In view of the above facts and inferences the judgment has been 
in favor of the Highlands group of crossings as the more defensible 
on geologic grounds as a route for the aqueduct line. Furthermore, 
in accord with the preferences already noted, the Storm King loca- 
tion is regarded as the most likely to give satisfactory results. 



Quality and condition of rock 

The rock of Storm King mountain and of Breakneck ridge at the 
Hudson river crossing is a very hard granite with a gneissoid 
structure of variable prominence. The color varies from grayish 
to light reddish and the structure is always coarse passing into 
pegmatite facies that occur as stringers or irregular veinlets. The 
grayish facies is of slightly finer grain and more gneissoid. Those 
portions that have been sheared are still darker. There are many 
joints at the surface running at various angles and an occasional 
slickensided surface. The mass is cut by several dikes of more 
basic rock (diorite) of widths varying from a few inches to 8 feet. 
These dikes are somewhat more closely jointed than the granite 
and consequently a little more readily attacked by the weather. But 
where protected they are equally substantial for underground work. 

The chief variation from ibis condition is where crushing or 
shearing has induced metamorphic changes. Wherever bed rock 
has been reached at this point and to such depths as workings have 
penetrated the rock is of this type. 

The work includes (a) four inclined drill holes from the river 
margin — two starting from the surface and two from chambers 
set off from shafts at a starting depth of about 200 feet, (b) several 
vertical holes in the river itself, and (c) two large working shafts 
20 x 20, one on either side of the river. 

These give all the data 1 known as to the condition at depth. From 
them it is apparent that crushing and shearing have been prominent. 
Many splendid specimens of crush breccia are thrown on the shaft 
dump. But its present condition at the depth involved is sound and 
durable. The fractures are rehealed. There has been a recombina- 
tion of constituents giving a new matrix of complex silicates among 
which epidote is the most characteristic, while simple decay is of 
little consequence. For strength and permanence the conditions 
could not well be improved. There is no reason to apprehend any 
change for the worse for the reason that the same tendencies must 
prevail at that depth throughout. It would appear therefore that 
faulting movements, or the existence of a fault zone of importance 
can not become a serious obstruction, because of the tendency to 

1 Since this paragraph was written four inclined diamond drill borings 
have .been made from chambers at depths of about 200 feet in the shafts. 
These have now penetrated the whole distance beneath the Hudson with 
very satisfactory results. 


lieal up the fractures and so make the rock about as substantial as 

It is noted elsewhere that faulting is common in this region, and 
that in a considerable portion of its lower course the Hudson prob- 
ably follows such structures. It is, however, wholly unnecessary 
to assume that its whole course is a fault line. WJiether or not 
there is a longitudinal fault zone of any prominence in the river at 
Storm King is unknown. There are several cross faults, both above 
and below this point, that give much clearer surface evidence of 
their presence. Fault zones have proven to be objectionable ground 
in many places along the aqueduct line, but elsewhere the data 
refer chiefly to situations favoring more ready underground cir- 
culation, i. e. at higher levels. In this particular case the rock in 
question lies below former ground water level within the belt of 
cementation rather than up in the belt of decay, and there is prob- 
ably no disintegrated rock from any cause. 



Topographic features of the southeastern margin of the Cats- 
kills, where the chief water supply is available, fixes the approxi- 
mate location and bounds of the principal reservoir. The accom- 
panying map, a portion of the western part of Rosendale quad- 
rangle, shows the situation. The part of the work involving the 
chief geological problem was the choice of the principal dam sites 
on the Esopus. This is known as the Ashokan dam. This part of 
the Catskill system belongs to the Reservoir Department under 
Mr Carlton E. Davis as department engineer. 

There were originally considered three sites: (i) at " Broadhead 
bridge," (2) at " Olive Bridge," (3) at " Cathedral gorge" or the 
" Tongore " site. Any one of these seemed possible from a topo- 
graphic standpoint. Later developments in regard to storage ca- 
pacity and engineering considerations finally reduced the practicable 
sites to two — the ''Olive Bridge" and the "Tongore." These 
were then explored thoroughly as an aid to determining whether 
or not there were favorable or unfavorable conditions at either 
location. Trenches were dug, shafts were sunk, wash holes were 
put down, and drill borings were made. The amount of such work 
done was sufficient to show the actual conditions both of the drift 
and bed rock and incidentally to throw some light on minor matters 
in geologic history. 

This discussion is essentially a summary of these data and a com- 
parison of the geologic conditions indicated by the explorations 1 of 
these two sites and a statement of some of the geologic character- 
istics of the area. 

1 General geologic conditions as shown by the explorations 

Bed rock is dark colored Devonic sandstone and shale, the 
Sherburne formation, lying almost horizontal, strongly jointed, 
plainly bedded, and of good quality for the foundation of the dam. 

At both locations the present Esopus flows in a postglacial gorge 

1 In this work of exploration a very efficient staff of engineers was engaged. 
Among those having very much to do with the features here discussed are 
Thaddeus Merriman, division engineer, J. S. Langthorn, division engineer 
and Sidney Clapp, assistant engineer. 




and there is a somewhat deeper huried channel a short distance to 
the north side. In each case this old channel bed rock is probably 
less fresh and substantial, due to former weathering, than the pres- 
ent exposed surfaces. 

In each case glacial deposits reach a thickness of more than 200 
feet within the narrow valley or gorge, especially along the north 
valley wall within the limits of the proposed dam. 

Special geological conditions. The factors in which there is 
most variation and which are of most significance in a comparative 
study are those belonging to the glacial drift deposits. In order to 
properly estimate the influence of some of these features it will 
be necessary to briefly consider the types of material represented 
at different places and the conditions under which they were 

Types of material. Till. Heavy bouldery till, mixed clay, 
sand, gravel, and boulders, is the most abundant type of material. 
It forms especially the chief surface material throughout the region, 
and is the surface type at both sites. It becomes at places quite 
sandy, but is almost everywhere good, impervious material because 
of its mixed character. 

Laminated till. At a few places, notably in the Beaver kill near 
its mouth, and in a trench above Olive Bridge, and in the " big 
ciugway " above West Shokan, strong lamination appears in heavy 
stony till as if laid down rapidly in comparatively quiet water such 
as the margin of a lake. This material is especially impervious. 

Gravel hillocks. A few small hillocks with morainic contour, 
indicating a dumping ground for some glacier on a small scale, 
occupy the flat immediately west of Browns Station. They were, 
at a very late stage, piled into the course of a former glacial stream 
whose delta deposits occupy the sandy bench above the 500 foot 
contour just north of Olive Bridge. 

Assorted gravel and sand. This material is abundantly developed 
just north of Olive Bridge. It seems to have formed a delta de- 
posit at the mouth of a glacial stream that emptied into the main 
valley at this point. The running water washed almost all of the 
clay and extremely fine material farther out, where they settled in 
the bottom of a small glacial lake that was at that time held in this 
upper portion of the Esopus valley. The dam that held in this body 
of water which reached above the 520 foot line stood near the 
proposed " Olive Bridge " dam site. The materials forming the 
dam were in part the glacial till that is now found on that site and 



in part the ice itself, which came in from the northeast, helping to 
complete the barrier. Into this lake the streams from the melting 
ice margins deposited their load of silt. This is well shown in the 
trenches cut across the terrace }i of a mile above Olive Bridge. 

A similar occurrence is seen at the cemetery near West Shokan. 

Laminated sand and clay. In all cases where silts were carried 
into the lake basin the finest materials were carried in suspension 
to greater distances from the margins, and slowly settled out in the 
form of alternate laminae of clay and fine sand. Each sandy layer 
represents a fresh supply of material and rapid precipitation of the 
comparatively heavy grains ; while each clay layer represents a 
period of greater quiet or decreased supply during which the finest 
particles settled to the bottom. A predominance of fine sand indi- 
cates either abnormal supply or proximity to the supply margin, 
while a predominance of clay represents either uniform and mod- 
erate supply or greater distance from the supply margin. 

These deposits are nearly impervious to water moving vertically, 
but much more pervious laterally and especially so in the most 
sandy portions forming the marginal facies. 

This type of deposit is to be seen at the surface at about the 
700 foot contour 2 miles north of Shokan, above the " big dugway," 
also in the trenches cut into the terrace at about the 500 foot con- 
tours Y\ of a mile north of Olive Bridge, and it is probable that 
this same type underlies the northern half of the " Tongore " site. 
The material marked " fine sand " at and below the 400 foot line 
on the accompanying " geologic section " G-H is judged to be of 
this type. 

Pebbly clays. These are developed to only limited extent and 
indicate probably floating ice in addition to the other methods of 

Gravel streaks and assorted pebble beds. Wherever water flowed 
with considerable current across the material either before or after 
deposition the finer particles were removed and only gravel and 
pebbles, too heavy to transport, were left behind. Some of these 
gravel beds were developed in the intervals of successive advances 
and retreat of the ice when for a time the lower valleys were unoc- 
cupied. In many places the succeeding advance of the ice would 
plow all these surface materials up again and mix them into the 
usual till ; but occasionally the oncoming glacier simply covered 
these deposits with its own till mantle, and they are preserved as 
records of these minor interglacial stages. Such behavior would be 



more likely to occur in the deeper channels. To this class of 
deposit belong some of the gravel beds of the " Tongore " site, 
notably that shown in one of the deep shafts. It is probable that 
the zones where the wash rig experienced a " loss of water " are 
most of them of this type. 


2 Summary of geologic history 

In preglacial time the Esopus valley was occupied by a stream 
of similar capacity to the present Esopus creek. Its channel lay 
to the north side of the narrow valley, having adjusted itself in 
conformity to the slight dips of the Hamilton sandstones and its 
principal joints. At the points under investigation this original 
channel is buried under several kinds of glacial deposits whose 
source of accumulation was chiefly from the north and northeast, 
blocking the stream channel and forcing the stream to the opposite 
(south) side. The direction of movement was favorable to the dam- 
ming of the Esopus creek valley and the deposits indicate that 
this occurred at several different times and at different elevations 
and that corresponding lake conditions occasionally prevailed. It 
is equally clear that there were intervals of retreat of the ice with 
attendant stream action and the development of gravel beds, fol- 
lowed by another ice advance, either obliterating the surface 
features or covering the previous deposits with another till layer. 
With each successive withdrawal the local streams found them- 
selves more or less completely out of place, and consequently their 
characteristic deposits formed in these intervals may be found in 
unlooked for places wholly inconsistent with present surface 

At the final withdrawal of the ice, Esopus creek found itself 
entrenched along the southern margin of the valley and has cut a 
postglacial rock gorge instead of removing the compact till from the 
original channel. But wherever only modified drift, either sand or 
clay, was the valley filling it scooped out great bends so that a large 
proportion of this type has been removed from the valley, and 
only the margins remain as terraces or covered beneath other pro- 
tecting deposits. 

" " . i 

3 Application to the choice of dam site 

o " Olive Bridge " site. The trenches and shafts together 
with surface exposures indicate that the glacial drift at the Olive 

in u> 
O Oj2^ 


'■3 v 
1-13 i f 

ns"o c rr 

5 « bo 
■r O 



^ a; 



3 * 

ft yi 

9 >> 

,2 m 
£ *° 

O ui 

2 S 

ft o 

P. "5 


eg ™ • 



Bridge site, at one stage in the glacial history, served as a natural 
dam and that water was successfully held above it to an elevation 
of 530 feet and perhaps more. 

CiTyOFNcwyvnit - 
.... _ _ .. — 

SmcltO" of Sif« On ConttrLm* 

On Act At 

Fig 16 Location of the Ashokan dam at Olive Bridge site and a geologic 
cross section. The small dots in the plan indicate exploratory borings. The 
section shows the rock profile indicating a preglacial channel of the Esopus. 
The present Esopus flows in a new postglacial ^channel at a higher elevation. 

The lowest materials in contact with bed rock are heavy stony 
till, laminated till and stony laminated clays — all good impervious 
material wherever exposed and tight upon bed rock. Sands and 
laminated clays are extensively developed immediately northward 
of the site and streaks of these deposits interlock to a limited extent 
with the till materials of the site itself, but they do not extend far 
and die out in wedges among the heavy deposits that characterize 
the southern slopes of the hill forming the northern terminus of 
the dam. These pervious streaks do not extend at any point con- 
tinuously through this hill and consequently as a whole the present 
barrier as it stands is practically impervious. The poorer materials 
(assorted gravels and sands) characterize the upstream side, and 
the better, more impervious materials (till and laminated boulder 
■clays) characterize the downstream side of the proposed Olive 


Bridge site. It is therefore advisable to locate any such structure 
as a dam at a point as far down stream on this site as other engi- 
neering factors permit. 

b " Tongore " site. At Tongore, bed rock is at least a hun- 
dred feet deeper than at Olive Bridge. In the deeper parts, below 
the 400 foot line the deposits as indicated by the wash borings [see 
sections] are interpreted as a fairly continuous succession of till, 
stratified sands and gravels, and laminated sands and clays belong- 
ing to two or three different stages of accumulation. Upon tins 
the heavy upper till was laid down. It is believed that the records 
fully support this view and that the stratified or laminated materials 
were accumulated at a time when a temporary dam existed at some 
point still farther down the Esopus valley. It is apparent further- 
more that the most porous zone is at the junction of the upper 
till and the lower stratified deposits and in part is represented by 
the assorted pebbles of stream wash — in general not far from the 
400 foot line. These middle zone deposits are believed to extend 
continuously through the drift ridge that forms the northern half 
of this site. As before noted, though rather impervious vertically, 

Fig. 17 Plan and geologic section at the Tongore site. The dots on the map indicate 
exploratory borings and the course of the buried channel of the preglacial Esopus creek 
is shown making a right angle bend to the north. The section shows the buried chan- 
nel, the new postglacial channel and the great accumulation of porous modified drift 
which is regarded as one important objection to this site for the dam. 

some of these deposits allow ready lateral movement of water. This 
is held to account for the rather persistent occurrence of springs 
or seepage along the creek bank at about this level both above and 



below the site. The great thickness of these laminated beds, in 
places a hundred feet or more, together with the abundance of sand 
in them, and the caving tendencies exhibited by them in one of the 
large shafts, indicates poor conditions for such a piece of work. 

The behavior of one of the test shafts throws some light on con- 
ditions within the drift deposits. At this place after sinking into 
the underlying gravel beds there was " no water " at first, but after 
going a few feet deeper there was an abundant flow, that did not 
rise much in the shaft. This case seems to support the following 

The gravels encountered do not form an isolated pocket or lens, 
else it would have carried water full from the first. It must be 
a fairly continuous porous zone with large feeding connections else 
it would run dry, and it must have an easy discharge else it would 
have risen above the level of the first gravels. Therefore it must 
be a rather well marked subterranean water passage or porous zone 
of considerable extent. Such conditions would make an impervious 
core wall to bed rock at this site a necessity and its construction a 
matter of considerable difficulty. At this site also because of the 
small cross section of the ridge, there is little chance for the inter- 
locking of layers or the blocking of the porous ones by a till barrier 
to check the lateral seepage, and there is no chance to move farther 
down stream to secure such conditions. 

4 Summary 

Because of the (a) higher bed rock throughout, and (b) the 
more uniform and impervious quality of drift deposits, and ( c) 
the more massive cross section of drift barrier for foundation, and 
(d) the perfectly tight contacts of till and bed rock, and (e) the 
limitation of the more porous materials to higher levels and (/) the 
glacial history connected with the development of all these parts. 
" Olive Bridge " is the preferable location for the proposed Asho- 
kan dam on Esopus creek. 



Probably no stone marketed in New York State is more exten- 
sively known than the " bluestone " of the Catskill region. But it 
is noted particularly for a special purpose, i. e. as flagstone, because 
of its capacity to part or cleave into thin slabs. These slabs are 
proven by experience to have remarkable weather resistance and 

Little attention has been given to the question of dimension stone 
— whether or not such blocks of as high quality as the flags could 
be obtained and where such quarries could be opened. 

There are several reasons for this situation. In the first place 
(i) the stone is of a dark color and has a dull appearance so that 
it is not fancied for the usual expensive structures where large sizes 
are used, also (2) the quarries are small, shallow, and are worked 
on a small scale by single individuals or groups of neighbors with 
few quarrying tools and no transportation facilities for large mate- 
rial, and in addition (3) considering the work and equipment neces- 
sary and the demand the flag industry was more profitable. 

Because of the large demands of the Ashokan dam where nearly 
a million cubic yards of heavy masonry construction are to be used 
an entirely new situation has developed. It is especially desirable 
that a rock capable of furnishing heavy dimension blocks should 
be discovered. The usual slab or flag type is unsuited to a consider- 
able part of this work. A study of the adjacent region therefore 
has been made and explorations along certain promising lines have 
been conducted to sufficient completeness to prove that a suitable 
stone can be furnished in large quantity. The characteristics of 
structure and occurrence as shown by this special study are given, 
together with some of the later exploratory data. 

Physiographic features 1 

All of the rock formations are sedimentary, chiefly sandstones 
and shales. They lie in alternating beds of variable thickness and 
are almost horizontal. The total thickness is many hundred feet so 

1 The principal argument of this discussion has been used in a previous 
article by the writer under the title " Quality of Bluestone in the vicinity 
of the Ashokan Dam" in the School of Mines Quarterly, v. 20, no. 2. 




that neither the hottom nor the top beds of the series are to be seen 
in this locality. 

The region is one of considerable relief representing preglacial 
erosion. The glacial drift mantle has modified it chiefly by obscur- 
ing some of the smaller irregularities of rock contour, and espe- 
cially by partially filling many of the stream gorges. Postglacial 
erosion has not completely reexcavated the old channels. But the 
contour of the uplands reflects the character of the bed rock with 
considerable success. The tendency of the more massive and coarse 
grained varieties of rock to resist weathering and erosion more suc- 
cessfully than the finer grained and more argillaceous or shaly 
facies is a general characteristic. Since these varieties form suc- 
cessive or alternating beds throughout the whole area, the result is 
an almost universal cliff -and-slope surface form. This bed rock 
topography is somewhat obscured but not wholly obliterated by 
glacial erosion and deposition. Therefore it may be used with con- 
fidence in locating or tracing the more durable beds since they 
almost invariably appear as a shelf or terrace with a steep margin 
toward the lower side and a gentle slope on the rising side. 

Structural features 

The rock types include bluish gray or greenish gray sandstones 
with almost horizontal bedding, and sometimes exhibiting cross- 
bedding structure, and compact very dark argillaceous shales. These 
two are of about equal prominence, but only the sandstone is of 
importance in the present discussion. Tts minute structure will be 
given in greater detail in the petrographic discussion. 

Jointing is common and persists in two sets nearly at right angles 
to each other — one striking northeastward and the other toward 
the northwest. In some of the best exposures, these joints are 
clear-cut and run 10 to 18 feet apart, dipping almost vertically. In 
the more massive beds there is very little small jointing, so that the 
character is especially favorable to large dimension work. 

But still more prominent structures are the partings which follow 
the bedding planes. These give the rock a decided tendency to 
cleave naturally into slabs, the uppermost exposed portion of almost 
every outcrop exhibiting this slab structure in more or less perfec- 
tion. So general is this structure at all horizons in the sandstones 
of the series that there can be no doubt of its connection with 
some original sedimentation character. Besides it is a potential 
factor in nearly all the beds even when not very apparent. The 



exposed places exhibit the character so prominently only because 
of the weathering effect, which develops the natural tendency. This 
general conclusion is borne out by the well known practice of quar- 
rymen of the district of splitting the larger blocks into slabs of the 
required thickness by wedges driven along certain streaks that are 
known as " reeds." A reeding quarry is one that has this capacity 
well developed, and it is this character in part that has made the 
" bluestone " or " flagstone " of New York an important factor in 
the production of the United States for a great many years. 

For large size dimension stone where great stress is involved it 
is evident that this structure would not be desirable. These definite 
planes of weakness reduce the general efficiency. A little observa- 
ti jn however shows that there are some outcrops and an occasional 
quarry where the more massive blocks do not split well. From the 
necessities of the industry these have been avoided or but meagerly 
developed. In some cases of this kind the sedimentation is of the 
cross-bedded type with somewhat interlocked laminae. If the grain 
is coarse such varieties resist splitting with great success. The 
thickness of such beds varies from a few feet to 25 feet or even 
more without prominent interbedding of shale layers. 

Stratigraphy. These are the sandstones, flags and shales 
known as the Hamilton, Sherburne and Oneonta formations belong- 
ing to the Devonic period. The strata of the immediate vicinity of 
this examination belong to the Sherburne subdivision, but no at- 
tempt to differentiate the formations was made. Structurally and 
petrographically the different formations are not distinguishable in 
this area. On the market the stone from either is known generally 
as " Hamilton flag " or " Milestone." 

Economic features 

There are hundreds of quarries in this general region. Nearly all 
are small, and are worked on a small scale without machinery. The 
product is almost wholly thin slabs of the flagstone type. This is 
supplemented by a small amount of somewhat more massive char- 
acter, dressed for window sills ; and a very limited output is of 
dimension stone of larger size. The general lack of suitable me- 
chanical devices and transportation facilities are the chief reasons 
for the limited output of the last named grades. 


The basis of this discussion is a microscopic examination of sev- 
eral thin sections made of the different types of rock from the 



quarries whose field geologic features give promise of encouraging 
results. The most characteristic variations are illustrated in the 
accompanying photomicrographs, plates 22, 23. 

Texture. The rock is granular, the individual grains varying 
from minute particles in the finer shale layers to three or four 
tenths of a millimeter in diameter in the coarser sandstone [pi. 23, 
lower figure]. The grains are seldom rounded. Jagged or frayed 
or elongate forms are the rule [pi. 23, upper figure]. There is no 
marked porosity. When the rock was first deposited as a sediment 
it probably had the usual large interstitial spaces of such rock type, 
but in this case some subsequent modification — an incipient meta- 
morphism — has largely obliterated the voids by the introduction or 
development of mineral matter of secondary origin. 

In general it is quite apparent that the average grain was orig- 
inally more rounded than its present representative. 

Mineralogy. The original minerals in order of abundance 
were the feldspars, quartz, and probably hornblende, biotite, and in 
much smaller amounts others of little apparent consequence in the 
present discussion. 

All of these have been more or less affected by subsequent 
changes. Quartz has suffered least of all, the chief modification 
being a greater angularity of form and an occasional interlocking 
tendency caused by secondary growth [pi. 22, lower figure]. 

Both orthoclase and plagioclase feldspars occur. The orthoclase 
grains, which originally made up more than half of the bulk of the 
coarser types of rock, have been in places profoundly altered [pi. 22, 
upper figure]. In many cases the identification of this mineral de- 
pends upon its association and the abundant remnants of character- 
istic structure and its normal secondary products. In the least 
affected grains satisfactory identification is not difficult. Even in 
the most modified representatives there is some preservation of 
structure indicating size of grain and proving the essentially gran- 
ular character of the rock. The plagioclase, although not abundant, 
is more readily detected than the orthoclase because it has been 
much less affected by the secondary changes. 

All original ferromagnesian constituents are wholly altered. There 
were some such constituents in the rock, as is plainly shown by the 
secondary products. Hornblende and biotite were probably both 

The secondary products, derived from the original feldspars and 
ferromagnesian constituents, include sericite, chlorite, calcite and 

Plate 22 

Photomicrograph of bluestone, x 25 diameters. The clearer grains are 
quartz and indicate the approximate size of other original constituents. 
In this case the alteration of the feldspars and ferromagnesian originals 
is so complete that their products form an indeterminable complex aggre- 
gate of closely interlocked granules, flakes, and fibers of extremely fine 

Photomicrograph of first grade medium grain bluestone, x 25 diameters. 
Taken to show angular and interlocking grains indicating secondary 
growth and a complete lack of reeding structure. The clear grains are 
quartz; the rest of the field is made up chiefly of secondary derivatives 
from the original feldspars and ferromagnesian minerals. 


quartz as the most important and abundant. Others probably occur 
that are less readily differentiated, and among them is kaolin. Occa- 
sionally a small amount of massive or granular pyrite occurs. There 
are traces of organic remains, especially plant stems, and the pyrite 
is most plentiful in association with those beds. 

It seems to be the secondary products largely that give the char- 
acteristic bluish or greenish color to this stone. Practically all of 
the iron freed by secondary changes from the ferromagnesian con- 
stituents has entered into new silicate compounds, especially with 
the chlorite, which are minutely distributed throughout the whole 
mass, giving it all a tinge of the characteristic color of these well 
known products. The same amount of iron in the oxid form 
would no doubt give as highly colored stone as any of the reds or 
browns of other familiar types of sandstone. But the tendency to 
form the sericite-chlorite-quartz aggregate in the rock has also an 
important bearing on its durability and strength. This is further 
discussed in a separate paragraph. 

Classification. It is clear that this type of bluestone is a sedi- 
mentary rock of medium grain, a sand rock or " renyte." Since 
the silicates are so predominant in the original composition it may 
be further identified as a sandstone or a " silicarenyte." But in 
view of the predominance of the feldspars it should be further 
designated as an arkose sandstone. And considering the extent to 
which it has been modified by the development of interstitial sili- 
cious products and the effect that this has had in perfecting the 
bond between the grains, the rock may be classified as an indurated 
arkose sandstone. 

Special structure. A study of the cause of reeding, or the 
tendency to split into slabs, led to the preparation of thin sections 
of this structure [pi. 23, upper figure]. It is apparent from them 
that the reed is strictly a rock structure and that the perfection of 
the capacity to split along these planes depends wholly upon the 
abundance and arrangement and size of the elongate and semifibrous 
grains and the presence of a more than usual amount of original 
fine or flaky material. Almost universally the reed streaks are 
darker in color and finer in grain than the average of the rest of the 

In part therefore it is an original character due to the assorting 
action of water during deposition, finer streaks alternating with 
coarser ones in accord with ordinary sedimentation processes. But, 
in addition to that, the subsequent changes that have affected the 
whole rock have occasionally accentuated the structure by a ten- 



dency of the whole rock to develop elongate or fibrous aggregates. 
It is probable therefore that the parting capacity is in places con- 
siderably increased by the very process that has produced just the 
reverse results in the more heterogeneous portions of the beds. 

Under a sufficient stress the rock will part most easily along the 
planes where this foliate or fibrous character is most persistent. 
Even in these cases, however, it may not indicate that the rock is 
essentially weak. It simply locates the most vulnerable point in 
the stone. In many quarries these streaks are so abundant that 
only thin slabs can be obtained — the disturbances of ordinary 
quarrying being sufficient to cause parting. The deeper portions of 
quarries are, however, much less subject to such behavior. In all 
cases the greater slab development of the exposed portion of the 
ledge is an ordinary weathering effect, by which the same results 
are obtained slowly and naturally and more perfectly than can be 
secured artificially on the fresh material of the same beds. The 
expansions and contractions of changes of temperature, together 
with the rupturing effects of freezing water caught in the pores, 
serve finally to weaken every part of the rock. In this process the 
prominent reed lines give way so much in advance of the rest of 
the rock that they develop into true rifts and separate slabs appear. 
It must be appreciated that these ledges have been exposed an im- 
mensely long time compared with the probable requirements of any 
engineering structure, and that this weathering tendency does not 
mean a speedy disintegration of the freshly quarried blocks. Still 
it is advisable to avoid as many sources of weakness as possible 
and one of the ways is to select ledges where the stone does not 
have a reeding tendency, or in which the reed lines are interlocked, 
or wavy, or interrupted. These requirements are most fully met in 
the coarser beds and especially those exhibiting some cross-bedding. 
Two local quarries meet these demands to a marked degree. 

Strength. The better qualities of bluestone have great 
strength. Even the reed lines are in many instances stronger and 
more durable than the regular quality of some other sandstones 
that are usually considered suitable building material. The secret 
of this exceptional strength lies in the modifications of texture that 
have resulted from the alteration and reconstruction of the mineral 
constituents. The breaking up of the orthoclase feldspar, and the 
accompanying changes in the ferromagnesian minerals, have fur- 
nished considerable secondary quartz, which has in part attached 
to the original quartz grains making them more angular and de- 

Plate 23 

Photomicrograph showing structure of the reeding quality of "blue- 
stone. ' Magnification 30 diameters. Taken to show tendency to paral- 
lelism of elongate grains. 

Photomicrograph of best grade coarse-grained Milestone. Taken to 
snow a quality m which the granular character is still well preserved. 
, 'i ! i"' SL ains , are quartz, the others are chiefly feldspars somewhat 
"Alined. ine close interlocking and the development of fibrous or 
trayecl structure and the bending or wrapping of some constituents are 
secondary effects. 


veloping an interlocking tendency [pi. 22, lower figure]. At the 
same time the fibrous sericitic and chloritic aggregates have developed 
to such extent as to fill most of the remaining pores, and in many 
cases the fibrous extensions have actually grown partly around the 
adjacent quartz grains [pi. 23, lower figure]. The effect has been 
to develop a silicious binding of unusual toughness. This combina- 
tion of changes has made a rock that is now remarkably well bound 
or interlocked for a sedimentary type. 

Durability. First-class stone of the grades indicated above 
would have as great durability as any stone in the market, except 
perhaps a true quartzite. With the exception of the almost neg- 
lectable quantities of pyrite, occasionally found, there is no con- 
stituent prominently susceptible to decay. The rock as a whole 
mineralogically is stable and its texture indicates unusual resist- 
ance to ordinary disintegrating agencies. 

General conclusions 

From the microscopic study it is clear that the variety of rock 
most fully meeting the demands of heavy exposed construction are 
the coarser beds and those freest from reed and shale. 

From the field study it is apparent that ledges of suitable char- 
acter occur occasionally and that at least three such are not far 
from the Olive Bridge site. 

From additional explorations it is certain that ledges of high 
grade rock occur, and that the grade varies rapidly in the same 
bed and that suitable material can be obtained in the immediate 
vicinity of the Ashokan dam. No doubt rock of equally high 
quality may be obtained at many other localities. 



Because of the fact that the hydraulic grade of the Catskill aque- 
duct as it approaches the Rondout valley is nearly 500 feet A. T., 
an elevation more than 300 feet above the lowest portions of the 
valley and more than 200 feet above very large areas of it, a total 
width of more than 4 miles being too low for unsupported con- 
struction of some kind, and because of the general policy of using 
the pressure tunnel system so as to deliver the water at a corre- 
sponding elevation on the east side of the valley, and further 
because of the very complicated geological features of the district 
this section has been the seat of very extensive and interesting 

Undoubtedly a greater number of obscure features occur here 
than on any other single section of the whole aqueduct line. Most 
of these features are readable from surface phenomena in general 
terms. In all cases the indications are plain enough to serve as a 
guide to well directed tests, but many points of critical importance 
can not be determined with sufficient detail and accuracy of posi- 
tion for such an engineering enterprise without systematic explora- 
tion. 1 The basis and results of this line of investigation which has 
occupied the greater part of two years are summarized and plotted 
in the following discussion and charts. The portion receiving 
special study is in the vicinity of High Falls. 

General geology 

Almost everywhere the surface is glacial drift. Where outcrops 
of bed rock occur they habitually present the unsymmetrical ridge 
appearance usually with a more or less sharply marked escarpment 
on one side and a gentle slope on the other. The strike of these 

1 These explorations belong to the Esopus division of the Northern 
Aqueduct Department. The earliest reconnaissance was done under the 
direction of James F. Sanborn, division engineer, who was subsequently 
assigned to geologic work over a considerable portion of the Aqueduct line. 
The development of exhaustive explorations and final construction on this 
division has been carried on under Lazarus White, division engineer, 
assisted by Thomas H. Hogan. The division has been recognized from 
the beginning as an important one and in many ways one of the most com- 
plex. Thomas C. Brown, now professor of geology in Middlcbury College, 
was employed for a year on this division during the later exploratory work. 




features is in general northeasterly and on the gentle slope is the 
westerly one. 

It is apparent at once that the valley bottom is a complex one 
and that its history has been somewhat obscured by the glacial 

Formations. The following distinct stratigraphic units are deter- 
minable in this valley every one of which will be cut by the tunnel 
beginning at the west side with the youngest formations : 


Hamilton and Marcellus flags and shales 700-+ 

Onondaga limestone t 200 

Esopus gritty shales 800-+ 

Port Ewen shaley limestone including the Oriskany transition 250-+ 

Becraft crystalline limestone 75 

New Scotland shaley limestone 100 

Coeymans limestone 75 

Manlius limestone including Rosendale, Cobleskill, and the cement 

beds • 100-+ 

Binnewater sandstone 50 

High Falls shale including small limestone layers 75 

Shawangunk conglomerate 250 to 350 

Hudson River slates — thickness unknown ; probably more than 2000 

Approximately 4775 

These occur in belts in succession more or less regularly from 
west to east. Most of the formations are quite uniform in the 
Rondout valley. The Shawangunk conglomerate is probably more 
variable than any other as shown by borings. Because of this 
general persistence of formation it is possible to estimate approxi- 
mately the depth at which any particular lower member lies if some 
starting point can be identified. [For detailed description of the 
formation, see pt 1] 

Structure. The principal irregularities are structural, rather 
than stratigraphic. The region on the west side of the valley, the 
margin of the Catskills, is but slightly disturbed and lies very flat, 
but the region on the east side, the Shawangunk mountain range 
and the cement district, has an extremely complicated structure. 
The Rondout valley, lying between them, is a transitional zone and 
passes from gentle dip slopes and folds in the westerly side to 
more frequent folds and thrust faults on the easterly side. In at 
least two thirds of the valley it would appear from surface evi- 
dence alone that the formations would dip uniformly westward, the 
only suspicion of additional complication being given by an occa- 



sional minor fold seen in the river gorge or an escarpment where 
the sedimentary character alone would hardly account for it [see 
pi. 24, High Falls]. Explorations have shown that the evidence of 
the minor structures is reliable and that disturbances occur at some 
places even to the extreme western margin. 

Physiography. In spite of the drift cover which obscures 
many original inequalities it is readily seen that the prevalence of 
the gentle westerly dip over most of the area, together with the 
succession of so many different beds of varying resistance to ero- 
sion, have allowed the development of a succession of long dip 
slopes and steep escarpments on a more pronounced scale than the 
present topography shows. It is clear that the Rondout is really a 
series of these unsymmetrical valleys. The principal large dip 
slopes are formed by the Shawangunk conglomerate and the Onon- 
daga limestone. In each case an original stream had adjusted its 
course fully to the structure and was shifting slowly by the sapping 
process to the west against the opposing edges of the overlying 
strata which form the bordering escarpment. One of these unsym- 
metrical valleys lies along the easterly base of the Hamilton escarp- 
ment and is continuous with the lower course of Esopus creek 
farther to the north. In the area under special study it is not 
occupied by a stream now but is filled with glacial drift so com- 
pletely that the original stream has been evicted. It is evident, 
however, from computations based upon the average dip of the 
slope carried to the base of the escarpment that the bed rock floor 
ought to be from 200 to 300 feet below the present surface in the 
deepest portion. Borings have proven this to be the case both 
along the present line near Kripplebush and also on the first trial 
line across the Esopus at Hurley. 

The same thing is true near High Falls in the center of the valley 
where Shawangunk conglomerate forms the dip slope and the 
escarpment is formed by the Helderberg limestones. In this case 
the drift filling is very deep also, and Rondout creek flows upon it 
quite independent of rock structure except where it has cut across 
the margin as at High Falls. 

In the eastern half of the valley the hard Shawangunk conglom- 
erate forms the chief rock floor and largely controls the contour by 
its own foldings and other displacements. Thus the Coxing kill 
tributary valley lies in a syncline of the conglomerate with occa- 
sional remnants of overlying beds as outliers adding some variety 
to the form. The Shawangunk mountains, as a physiographic 



feature, owe their present elevation chiefly to the resistance of this 
conglomerate which serves as a protective member among the 

On the west side, the foothills of the Catskills form a part of 
the cuesta developed by the erosion of Paleozoic sediments, the 
inface coinciding with the escarpment along the lower Esopus and 
Rondout valleys at this point. 

It is certain therefore that the drainage of the Rondout valley 
before the Ice age differed materially from the present lines. A 
stream, probably the original Rondout, followed near the western 
margin of the valley and joined the Esopus as it emerged from the 
Hamilton escarpment to turn northeast. Another which had cut 
somewhat deeper occupied the central portion of the valley and 
probably joined the Esopus at some point farther north — its lower 
course is not explored. 

Practical questions 

The chief practical questions to be given as full answers as pos- 
sible are: 

1 At what depth must the aqueduct tunnel be placed in order 
to be everywhere in substantial bed rock with sufficient cover to be 
safe ? 

2 Where are the most critical places — those whose geologic 
characters are such as to demand exploration ? And at the same 
time which sections may be safely left without testing? 

3 What is the rock structure and condition? And are. there rea- 
sons for believing that the tunnel plan is not feasible at this point. 
If so, where can a better one be found? 

4 What is the character of underground circulation of water? 

5 What formations will be cut at the different points and which 
should be favored or avoided wherever possible? 

From the fact that the present Rondout flows across solid ledges 
at High Falls and at Rosendale from ioo to 200 feet above the 
known rock floor of the preglacial gorge where explored it is clear 
that the present course is entirely different from the original. The 
Coxing kill, the third and most easterly of these streams is not so 
much disturbed although it also is shifted. 

It is worth noting that the streams of this valley together with 
the lower Esopus and the Wallkill river have become so completely 
adjusted to the rock structure that they all flow up the larger 
Hudson valley, of which all form a part, and join the master stream 



at an obtuse instead of the usual acute angle. They are essentially 
retrograde streams. 

Explorations. Systematic explorations and tests are repre- 
sented chiefly by drill borings through drift into the rx>ck floor. 
These were supplemented by two test tunnels for working character 
of material and a series of tests on the behavior of certain of the 
drill holes, together with other tests on material. The results are 
embodied in the accompanying cross sections and the additional 
discussion of special features. 

Detail of local sections 

Kripplebush section. This from the first was regarded as one 
of the critical sections because of the buried gorge along the base 
of the Hamilton escarpment and because of the doubt as to the 
behavior of the Onondaga limestone. On the accompanying section 
the borings are plotted and the structure as now interpreted is 
indicated. The dip slope formed by the Onondaga limestone is 
covered by 200 to 250 feet of drift, mostly modified drift. The 
strong valley character of the rock floor is almost wholly obscured 
by the glacial deposits and the present brook, an insignificant stream 
compared to the preglacial one, occupies a position above the escarp- 
ment instead of above the old channel. 

After a couple of the central holes were finished, it became appar- 
ent that the structure is not nearly so simple at this point as the 
general surface features would lead one to expect. It was clear 
that a simple dip such as was proven to prevail on the dip slope 
would not account for the much greater depth attained by it in 
the vicinity of station 500. The discovery of this additional feature 
raised two questions: (1) Is the structure a flexure or is it a fault, 
and if a fault whether normal or thrust, and (2) what is the prob- 
able effect of this structure on the position and depth of the pre- 
glacial gorge ? 

The habit of the district immediately east of the valley would 
support the theory of a thrust fault. The nature of the immediate 
area would suggest a simple flexure while it is manifestly possible 
that a normal fault could easily occur. Later explorations 1 have 

1 Since the above was written the tunnel has been completed through the 
Kripplebush section. Although faulting is indicated by the borings and 
actual occurrence of the beds it is very difficult to find the fault. A part 
of the displacement is accomplished by the steepening of the dip but this 
will not account for more than half of it. 


tested this zone so well that it is practically certain that the feature 
must he regarded as a fault of some type with a displacement 
of nearly 200 feet. The striking physiographic feature is the 
development and preservation of the escarpment on the downthrow 
side. This occurrence is certainly a very unusual case in that 
regard [see fig. 19]. 

Because of the intention to construct the tunnel deep enough in 
bed rock to reach safe rock conditions the question of depth of 
buried gorge becomes an important one. As soon as it was dis- 
covered that a fault existed there the problem became of sufficient 
prominence to demand more detailed exploration. If the faulting 
is accompanied by a broken zone in condition favorable to more 
ready erosion, it would be possible that the original stream in work- 
ing down this dip slope might become entrenched in the fault zone 
and at that point begin to cut a narrow gorge instead of continuing 
the sapping process. In fact, it would undoubtedly do this very 
thing if there is such a crushed zone of any consequence and if the 
erosion process were allowed to continue long after reaching this 
critical point. 

As a matter of fact explorations have shown that there is a thin 
layer of Hamilton shales still remaining on the Onondaga and the 
deepest point found is on the Hamilton shales side. These facts 
in connection with the failure to find any deep notch indicate that 
there is probably no zone of much greater weakness than the shale 
member itself. It is reasonable to conclude that the rock floor can 
be safely regarded as not much lower than 88 feet A. T. and that the 
rock condition is not especially bad for tunnel construction 1 even in 
the fault zone. 

Rondout creek section. This is the central portion of the 
valley including the depression occupied by the present Rondout 
and the exposed edges of the series of shales and Helderberg lime- 
stone. The repetition of the dip slope and escarpment, together 
with the heavy drift filling and the occurrence of so many forma- 
tions together make this an important section. All formations from 
the Shawangunk conglomerate to the Port Ewen shaly limestone 
occur at this point, and although there is little outward evidence of 
disturbance it is certain that whatever difficulty is to be found in 
this variable series is likely to be met here. It is therefore a sec- 
tion that requires exploration both for depth of preglacial channel 
and for quality of rock. 

In construction this ground has proven to be good and sound throughout. 



All of the formations dip westward wherever exposed, but the 
dips vary somewhat, nearly all being of low angle. Occasional 
minor inequalities of the nature of small rolls may be seen, as, for 
example, the small fold in the gorge at High Falls [see pi. 24]. 

Explorations have shown, as indicated on the accompanying cross 
section [fig. 20], that there is a deeper buried gorge here than at 
Kripplebush. The deepest point discovered is a few feet below tide 
level. The escarpment is steep and is formed by the Coeymans and 
New Scotland formations. The dip slope is Shawangunk conglom- 
erate, High Falls shale and Binnewater sandstone, with the Manlius 
limestone forming the floor. 

Identification of the drill cores which penetrate the limestone 
indicate that the dip slope is reversed on the west side of the gorge 
and that the stream had really reached about the axis of the trough. 
A discrepancy in thicknesses and depths in hole no. 34 by which it 
appeared that the Coeymans formation was almost twice as thick 
as usual and that it contained a broken or crushed zone leads to 
the interpretation that there is a small thrust fault here which re- 
peats the formation as shown on the accompanying cross section. 

Instead of a uniform westerly dip of all formations from the 
Rondout westward it is proven that minor anticlinal rolls and even 
thrust faults, as in this case, or such faults as in the Kripplebush 
case are not to be excluded. 

This structural relation has a direct bearing upon the question of 
the thickness of the Esopus shales. The Esopus is certainly not so 
thick as would otherwise be supposed, by 200 or 300 feet at the 
least. The true thickness is still an unknown quantity (estimated 
at 800 feet). 

It is clear that the aqueduct tunnel will have to be constructed a 
considerable depth below sea level at this section, probably not less 
than minus 150 feet, 1 even if the character of the formations be 

But the character or quality of these formations in view of their 
structural relation constitutes the chief problem. Because of the 
fact that every structure reaches the surface and eventually dips 
gently to the west in such manner as to encourage water circulation, 
their water-carrying capacity or general porosity becomes of great 
importance. A great capacity is all the more serious because of 
the heavy drift cover within the abandoned gorge, on top of which 

1 This portion of the tunnel and its continuation south to the Shawangunk 
range has been constructed at 250 feet below sea level. 



the stream flows and which constitutes essentially an unlimited 
storage reservoir to feed underground circulation. This is all the 
mor-e true if crush zones are extensively developed as accompani- 
ments of the faulting. 

In general as to perviousness the indications are somewhat ob- 
scure. But the data now obtained seem to prove that all the for- 
mations except the Binnewater sandstone and the High Falls shale 
are compact and fairly impervious along the bedding lines. Only 
where crevices have formed or where crushing occurs is there 
likely to be heavy circulation. This is all the more important since 
so many of the beds are limestones known to be readily soluble in 
circulating water. One of these limestones, the Manlius, exhibits 
occasional large open solution joints at the surface — so large that 
a surface stream disappears entirely at the so called " Pompey's 
cave " and joins the subterranean circulation. But such caves are 
probably limited to the surface. 

It is near this point, however, that one of the earlier borings at 
one side of the present line discovered very soft ground at a depth 
of about sea level, i. e. over 200 feet below the present surface, 
which shows that similar conditions prevail at certain points to 
great depth. 

Pumping tests made on hole no. 32 in an attempt to establish 
some data on the inflow of water gave very interesting results. 
These tests were very thorough. It was proven that the water was 
supplied in apparently inexhaustible quantity at maximum pumping 
capacity, which was ninety gallons per minute. Futhermore, the 
chief inflow seemed to be from the Binnewater and High Falls 
formations as was to be expected. Whether a crush zone allowing 
free circulation is furnishing a portion of this supply or whether 
the whole inflow represents the normal porosity condition of these 
formations is not yet proven. 1 

Other porosity tests have been made in such way as to locate 
and measure this factor [see later discussion]. Hole no. 10 shows 
an artesian overflow that comes from the Binnewater sandstone. 
A working shaft has been put down also in the vicinity of hole 
no. 32 and at the same depth found an enormous inflow of water 
which drowned out operations for a time. The lateral supply in 
this case has been reduced by introducing a thin cement grouting 
through holes bored in the surrounding rock from the surface. 

Holes no. 12 and no. 14 also show an artesian flow, but both are 

1 In construction the Binnewater sandstone has been found very wet. 



"shallow holes and the supply comes from near the contact between 
High Falls shale and Shawangunk conglomerate. 
/ It is certain from these observations and tests therefore that the 
Binnewater sandstone and High Falls shale are more porous than 
the other formations, and because of the serious difficulties arising 
from so heavy inflow of water from them the tunnel grade should 
be shifted so as to avoid these formations as much as possible. A 
comparison of the accompanying cross section, which is drawn to 
scale [fig. 20J , will show that a tunnel on one level would neces- 
sarily run for a long distance in these beds because of the gentle 
syncline. Furthermore, they lie at about the depth that would 
otherwise be a safe depth below the buried gorge. But a tunnel 
with a step-down, i. e. one run at two different levels could avoid 
most of this poor ground. By approaching at a level of about — 50 
feet or — 100 feet in the limestone beds to station 600 (hole no. 34), 
then stepping down to — 250 feet, the line in a very short distance 
crosses these two porous formations and enters the Shawangunk 
conglomerate which is more substantial, and, all things considered, 
one that seems most advantageous for successful construction. It 
will have to maintain a head of more than 700 feet as the difference 
between hydraulic grade and the tunnel level in this section. Under 
these conditions rock quality and condition are of greatest impor- 
tance and there is no doubt about the advisability of avoiding the 
poorest formations in some such manner. 

Coxing kill section. On the line of exploration the Coxing 
kill flows over Shawangunk conglomerate and High Falls shale. 
Both dip plainly eastward, and a hole no. 1 1 located on the east 
side of the brook penetrates about 70 feet of drift and shale. But 
only a hundred feet to the east Shawangunk conglomerate outcrops 
at the surface dipping the same way. It is certain therefore that 
a fault occurs here. The dip of the fault plane is indeterminate 
from the surface, but the relations and surroundings indicate a 
fault of the thrust type. 

Later explorations indicate that the fault plane is rather flat 
[see cross section fig. 21] so that the shales are repeated above 
and below a tongue of conglomerate. Boring no. 1 1 has also an 
artesian flow of considerable volume coming from near the bottom 
of the conglomerate. It is a mineral water. 

The chief importance of this section as a problem in applied 
geology lies in the influence of the fault and the maximum de- 
pression of the conglomerate. If the tunnel, which enters Hud- 
son River slates at the Rondout creek section at — 250 feet can 
be kept within that formation throughout the rest of its course, 

i 3 6 


there is no doubt that an advantage will be gained both in the 
greater imperviousness of the rock and the greater case of pene- 
tration. Wherever the conglomerate is undisturbed it is perfectly 
good, but where broken the crevices are but imperfectly healed 
and circulation is unhindered. It would therefore be desirable to 
know whether at — 250 feet the whole of the downward wedge of 
Shawangunk could be avoided. The borings indicate a thickness 
of Shawangunk of 345 feet in hole no. 11 where it is cut at a 
small angle, and a thickness of 409 feet in hole no. 36 where it prob- 
ably lies pretty flat. This greater thickness together with the 

finding of crushed rock at about the — 100 foot level leads to the 
conclusion that the formation is overthickened here by the thrust 
fault to the extent probably of about 75 feet. The true thick- 
ness of the formation at this point is doubtless more nearly 300 
feet than either of the figures obtained directly from the two 
holes. If this interpretation is used as the basis of plotting a 
cross section [sec accompanying cross section] it is apparent that 
the conglomerate should not be expected to extend more than a 
few hundred feet east of hole no. 3 6 and it probably does not reach 
a much greater depth than the — 236 feet represented as its base- 
in that boring. 1 

1 Construction of the tunnel has progressed far enough through this sec- 
tion to prove that the Shawangunk formation' does not reach much lower. 
It forms the roof of the tunnel for some considerable distance but does not 
come down into the tunnel more than a foot or two. 



Shawangunk overthrust. At the extreme eastern side of the 
Rondout valley near the point where the surface reaches hydraulic 
grade again, the surface outcrops pass from High Falls shale to 
Shawangunk conglomerate to Hudson River shale in the normal 
order but with entirely too small an area of conglomerate consid- 
ering the character of the formations. The higher ground is all 
Hudson River in the vicinity, and there is abundant evidence of 
crushing and disturbance. It is evident that a thrust fault is again 
encountered here, one of sufficient throw to bring the Hudson River 
slates above the Shawangunk conglomerate — probably a lateral 
displacement of very great extent. Explorations have fully proven 
the existence of this fault. The accompanying diagram shows a 
cross section as now outlined by complete penetration of two 

Two trial tunnels were run to test working quality of Hudson 
River slates compared to Shawangunk conglomerate at this locality. 
Both are within the influence of the fault zone. Both are there- 
fore more broken than the normal with the result that the Hudson 
River slates probably show poorer condition than usual and more 
troublesome working, while Shawangunk conglomerate probably 
shows easier working than usual. It is believed that normally the 
two rocks would present a greater difference than was found in 
this test. 

Special features 
Several questions, some of which have a practical bearing, have 
been raised as separate features during the exploration of the 
Rondout valley. 

Caves. One of these is in regard to the possible existence of 
underground caverns. This was given a special prominence early 
in the work by the experience of one of the drills. After pene- 
trating the limestone series near High Falls to a depth of over 
200 feet, the drill seemed to leave the rock and enter a space 
allowing the rods to drop 28 feet before being arrested by solid 
material. The further attempt to work in this hole resulted in 
the breaking of the rod doAvn at this point and the subsequent 
failure to recover the diamond bit which is still in the bottom of 
the hole. The question is as to the meaning of this occurrence. 
Is it a cavern? 

" Pompey's cave " has been referred to in an earlier paragraph. 
This is clearly not much of a cave. It is essentially an enlarged 
joint or series of joints by solution along the bed of a surface 



stream to such extent that the stream normally at present has be- 
come subterranean. It is the writer's opinion that the case en- 
countered by the drill boring is similar. The apparent cavern is 
probably a slightly enlarged joint along a line of somewhat abun- 
dant underground circulation and perhaps associated with some 
crush zone developed by the small faulting known to occur in this 
immediate vicinity. It is probably not entirely empty but contains 
residuary clay, and in all likelihood is very narrow and not exactly 
vertical, so that the drill rods were bent out of their normal course 
and wedged into the lower part of the crevice. Smaller spaces 
of this sort were encountered at a few other points. 1 

These occurrences seem to indicate that the limestone beds yield 
rather readily to solution by underground water, and that this cir- 
culation has been at one time active to at least 50 feet below pres- 
ent sea level. With present ground water level nearly 200 feet 
above sea level it is extremely unlikely that any such action is 
going on at so great depth. The occurrence is therefore strongly 
corroborative of former greater continental elevation when the 
deep stream gorges, now buried, were being made. These deeper 
caverns or solution joints probably date from that epoch. 

Imperviousness and insolubility. The question of impervious- 
ness and closely associated with it that of solubility, is of great 
practical importance in this particular work. The immense pres- 
sure under which the tunnel will be placed in crossing this valley 
makes it impossible to construct a water-tight lining. Everywhere 
much depends upon the rock walls to help hold the water from 
sjeriotis iloss. Wherever the rock is fairly impervious except 
occasional crevices or joints they can be grouted and safeguarded 
satisfactorily. But where a formation is of general porosity this 
can not be so successfully done. Even more difficult to handle is 
the rock wall which is soluble and which therefore with enforced 
seepage may tend to become progressively more porous. That this 
consideration is not wholly theoretical is shown very forcibly by the 
Thirlmere aqueduct of the Manchester (England) Waterworks. 
In that case a 3 mile section was built through limestone country 
using the same local limestone for concrete aggregate. Although 

1 In constructing the tunnel several clay-filled spaces have been discovered 
in the same vicinity at elevation — 100. One of these extended vertically with 
a width of 1 to 2 feet and from it a great mass of mud ran into the tunnel. 
At one point it was connected with a horizontal space of the same kind 
extending 15 feet. It can be seen that the original crevices have been en- 
larged by water and that they were originally formed during faulting. 



this concrete was mixed as rich as 1 part cement to 5 parts aggre- 
gate and the work was well done, excessive leakage reaching a total 
of 1,250,000 imperial gallons per day was developed within a year. 
It was found that the limestone fragments of the aggregate were 
corroded forming holes through the lining of the aqueduct and that 
these holes actually enlarged outward. All this was done under cut 
and cover conditions with not more than a 6 or 7 foot head on the 
bottom of the aqueduct. 

In the Rondout valley, the aqueduct will cut no less than 6 lime- 
stone beds in all cases under great pressure. This fact will in all 
probability tend to increase the action. But, of course, some of the 
beds may not yield so readily to solution. Tests made thus far, 
however, indicate that all are attacked in water. Considering these 
facts it seems desirable, so far as possible, to avoid the limestone 
beds wherever rock of greater resistance to solution can be reached, 
and further it is equally desirable to use a more resistant rock for 
the lining concrete. So long, however, as the formation is not very 
pervious so that a new circulation could not be established by the 
escaping water there would be little harmful effect. 

An average of five analyses of the Thirlmere limestone, different 
varieties of the same formation, gives the following: 

Insoluble silicious matter 2.772% 

Alumina and iron oxid Al 2 3 +Fe 2 3 0.276 

Lime, CaO 53.676 

Magnesia, MgO .390 

Carbonic anhydrid, COs 42.248 

Total 99.362 

Estimated calcium carbonate, CaC0 3 — 95.85,^ 
The limestone is fossiliferous. 

Suitable analysis of the limestones of the Rondout valley are not 
recorded in sufficient numbers. But these are a few, as given below. 


At Hudson 

At Rondout At Wilbur (av'ge of 2) 

Si0 2 

A1 2 2 

Fe 2 O s 



co 3 . 

3.87% 7.10% 1.865% 

1.07 2.50 .818 

1,34 1-65 1-185 

54-H 45-32 51-375 

tr tr 2.870 

40.60 39.10 40-795 

Total 100.99 

Corresponding to total calcium carbonate.. 96.62 






This is a limestone that in composition and structure at the 
Rondout valley is apparently not very different in quality from the 
Thirlmere rock. Analyses of the cement rock show less similarity 
but observations indicate that it is also attacked. 

It is probable from all these facts that the shales and conglomer- 
ates are better quality of wall than the limestones. 

A very acute observation along this line by Dr Thomas C. Brown 
while employed on the staff of the Board of Water Supply is of 
special interest. In studying local conditions he noticed that the 
limestone blocks used in building the old Delaware and Hudson 
(D. & H.) canal showed the effect of contact with the water. The 
best place for measurable data seemed to be around the old locks 
where squared and evenly trimmed blocks had been used. These 
were, during the years of its use, from 1825 (approximately 35 to 
40 years) subject to the action of water flowing or standing in 
direct contact. The coigns of the locks, which were without doubt 
freshly and well cut when laid, are now etched till the fossils and 
other cherty constituents stand out from one eighth to one half 
inch beyond the general block surface, and in some cases the pits 
are an inch deep. That this etching is due to the water rather 
than to exposure to weather is shown by the lack of such extensive 
action on blocks used in houses and exposed a much longer time. 
Blocks representing the Manlius and Coeymans were identified. 
But there is no reasonable doubt that others would be similarly 
affected. On some it would be less easily detected. 

On account of the disturbances another factor is introduced. 
Rocks which readily heal their fractures are likely to furnish better 
ground, i. e. more free from water circulation especially, than rocks 
more brittle and slow to heal. Therefore in this district the shales 
and slates such as the Hudson River series and the Esopus and 
Hamilton shales are the best ground, while the Binnewater sand- 
stone is the poorest. 

Cross sections. Probably in no region of like extent is it 
possible to> construct a geologic cross section of so many complex 
features so accurately as can now be done of the Rondout valley 
along the aqueduct line. The section is known or can be computed 
to a total depth below the surface of 1000 feet, including 12 dis- 
tinct formations, so closely that any bed or contact can be located 
within a few feet at any point throughout a total distance of over 
4 miles. 

The accompanying cross section contains as much of this data 
as is now available [fig. 22]. 


! * 


s 5 

* it 

1 , {•] i 


$ 1 Mill 




Vf- — i i- - 

/of m 

6i I 






I 4 I 

Rondout siphon statistics 

1 Total borings on the siphon line. Three different boring 
equipments have been used owned by different parties and records 
have heen kept so that the work of each can be followed or com- 
pared with the others. 

On this division the Board of Water Supply owned and operated 
one machine with their own men, another equipment was owned 
and operated by C. H. McCarthy, while a third which finally did 
a majority of the work, belonged to Sprague & Henwood, Con- 
tractors, of Scranton, Pa. 

The totals of different general types of material penetrated by 
these machines are as follows : 




Per cent 



feet of 

of core 





a B. W. S. Equipment 

1740. 5 




b Sprague & Henwood 









The average saving of core by all machines, cutting all kinds of 
bed rock was 75.96^ 

2 Core recovery from various strata. So nearly as can be 
done the strata represented in the drill cores have been identified 
and summarized as to total penetration and core saving. The core 
saving is a factor of prime importance in judging of the quality of 
rock and its freedom from disturbance. The following items are 
gathered from a study of the whole series. 

a Holes 6, 10, 12, 13, 15, 17, 18, 21, 22 and 25 penetrate Helder- 
berg limestone, a total combined depth of 1096 feet. Individual 
holes vary in core saving from 39.3;/ (no. 13) to 95.3$ (no. 15). 
The average core saving is 78.19$. 

b Holes 8 and 9 are in Onondaga limestone with a total pene- 
tration of 197 feet. The core saving varies from 56.2^ to 92.8%' 
with an average of 74.5^. 

c Holes 11, 19, 20, 23, 24, 27 penetrate Hudson River shale and 
together represent a total of 696.5 feet. The core saving varies 
from 16.6$ to 89^, with an average of 42.1$. 

d Holes 6, 10, 11, 12, 14, 16 and 20 cut High Falls shales to a 
combined total of 410 feet. The saving varies in different holes 
from 17$ to 83.3^, with an average core saving of 44.5$. 

e Holes 8 and 26 penetrate Esopus shale and penetrate 76 feet. 



The core saving varies from 73$ to 84.6^, making an average of 

/ Holes 10, 11, 12, 14, 16, 19, 20, 23, 24 and 27 penetrate Shawan- 
gunk conglomerate a total of 1356.5 feet. Core saving varies in 
different holes from 33.3$ to 100$. The average recovery is 60.52^'. 

y Holes 6, 10, 12, 15 and 16 cut Binnewater sandstone. The 
total penetration is 205 feet. The range of core saving is from 
30.6^ to 74.7^, with an average of 56$. 

h Holes 7 and 9 cut Hamilton shales to a total amount of 65 feet. 
The range of saving is 70$ to 81. 8$, with an average of 75.9$. 

3 Artesian flows. Several of the borings struck artesian flow 
of water. The fact that the sources of this flow are not the same 
has led to a tabulation of these data. 


,' Static Flow encountered 

Hole Size in head gallons at elevation 

no. inches in feet Minute Day Feet Strata 

10 i 18 30 43 200 — 109 ....Binnewater sandstone 

11 1 10 10 14400 — 60 . . . . Shawangunk conglomerate 

12 y% 1 ■ — 24 ....High Falls shales 

14 Wa +90 

20 Y& 7-5 i° 14400 +108 ... .Shawangunk conglomerate 

23 2 — 5 . . . . 

31 2 +158 .... 

39 1J/2 +112 . . . .Helderberg limestone 

S'NE Y% 12.4 .... 432 +203 Hamilton shale (possibly 


Pumping experiments and porosity tests 

Systematic tests have been made for flow of water, behavior of 
ground water and porosity of rock on certain of the Rondout ex- 
ploratory holes under the direction of Mr L. White, division engi- 
neer. A summary of these tests has been furnished by him from 
which is quoted the following: 

In addition to determining the location and thickness of the beds 
and the general character and condition of the rock from inspection 
of the cores, serious attempts were made to determine the relative 
porosity and water-bearing quality of the rocks encountered for 
the following reasons. (1) To determine the probable leakage from 
the siphon when in operation. (2) To determine the probable 
amount of water to be handled in construction. These experiments 
were divided into three classes: (1) Observation of flow from cer- 



tain drill holes which showed sustained flow of water. (2) Pres- 
sure tests in which water was pumped into holes which had been 
sealed off and pressure and leakage noted. (3) Pumping tests in 
which water was pumped from 4 inch drill holes by means of deep 
well pump of the type used in oil wells, and fall of ground water 
during pumping and subsequent rise after cessation of pumping 
noted. A description of the first two and the results obtained from 
them follows : 

A substantial flow of water was observed from the following 
holes : 

11/17: 50 gallons per minute through 2^/2 inch pipe, static head 
10 feet 

10/17: 30 gallons per minute through V/z inch pipe, static head 
18 feet 

20/17: 10 gallons per minute through }i inch pipe, static head 
7.5 feet 

The static head was observed by adding on lengths of pipe until 
the water ceased to flow over. It will be noticed in the case 
of hole no. 10 that the flow from the i J /> inch pipe is not that due 
to static head of 18 feet, but that due to a head of only y 2 foot. In 
other words the friction head is about 17.5 feet, and the velocity 
head only ^2 foot. This' same condition holds true of the other 
holes from which a flow was obtained. This would seem to indicate 
that the amount of water is not very great but that it is under con- 
siderable pressure. It is believed that this pressure is caused by 

A slight flow was observed from the following holes: 12/17, 
H/17, 23/17, 31/44, 39/ 22 , and 5/NE. 

The flow from most of these holes has ceased since the pipe used 
in boring was withdrawn. There is still some flow from the follow- 
ing holes: 11/17, 20/17, 25/17 and 5/NE. 

The flow from hole 11/17 is constant at about 10 gallons per 
minute. The others are too small to be measured. It will be noted 
that the only substantial flows encountered were from the High 
Falls shale, Binnewater sandstone and Shawangunk grit, and that 
it was possible to force water into these rocks in greater quantities 
and at a less pressure than in the other shales and limestones. 

Porosity tests. The method of making these tests was as 
follows : 

Wash pipe ecmipped with a device for sealing the hole was 
lowered to the desired elevation. The seal consisted of alternate 
layers of rubber and wood around the pir>e preventing the water 
from escaping between the walls of the hole and the pipe. Water 
was then pumped in and pressure and leakage noted. 

The result of the pressure tests was to show in a general way: 
( 1) That the pressure increased with the depth of seal. (2) Thai 
the leakage decreased with the depth of seal. (3) The maximum 
pressure in the grit was T40 pounds to the square inch and minimum 



leakage was 5 gallons per minute. (4) In the Hamilton shales a 
pressure of 300 pounds to the square inch with very little leakage 
was obtained. 

The unknown factors are too many and too great to make any 
reliable deductions from these experiments. 






so MAV 






'El- - JAW . • 



1 29- MAR 27 









Cn/el APR 
APR 9-I + .P* 

MPtO 439 ( 


CtO I 7 





APR IS-**-/ 


1.000 fAU 

ICED <h 6' 



PUrtPtO 64 
LL/El R£ 

> OOO ^ al-- IN 

17MR*/ ' 

^ — 1 

punpino A* 






FROM i> A.M - 

jnPiMc 5 a 
feP.M., WAT 

M .- 5 • JO r^, 

: a ROSt 2 

Fig. .23 Curve 'showing fall of ground water level while pumping_from boring 34 

Pumping experiments were carried on in holes 32/22 as follows: 
The apparatus used was a deep well pump of the type used in oil 
wells. The holes were of an inside diameter of 4*4 inches and 
were cased to the bottom. A 3/2 inch working barrel was then 
lowered to the bottom of a line of wooden sucker rods. The stroke 
was 44 inches and the nominal capacity of pump at 38 strokes per 
minute was 60 gallons per minute or 86,400 gallons per day. The 
power was obtained from a 40 horsepower boiler and 35 horse- 
power engine belted to a 10 foot band wheel which was connected 
to a 26 foot walking beam. In hole 32/22 at station 607 + 50 the 
average discharge at 38 strokes per minute was 90 gallons per 
minute or 129,600 per day. The experiment was continued for 15 
days and the total amount of water pumped was 1,071,000 gallons. 
The ground water level was not lowered. It will be noticed that 
the discharge at this point was 50$ in excess of the theoretical 
capacity of the pump. This was caused by the presence of gas, the 
effect of which seemed to be increased by the churning action of 
the pump. This may also explain the failure to lower the ground 

The experiment at hole 34/22 was similar in character. The 
upper 230 feet of this hole had an interior diameter of 4] '4 inches 



and the bottom 274 feet a diameter of only 3^4 inches. At first a 
2^4 inch working barrel was used to pump from the bottom and a 
discharge at 32 strokes per minute averaged 24 gallons per minute 
or 34,500 gallons per day. This was continued for about 15 days 
and the total quantity pumped was 490,000 gallons. The ground 
water level was lowered 17 feet at hole 34 and 4 feet at hole 32, 
750 feet away. 

The 3/4 inch pump was then let down to a depth of 200 feet 
with a 2*/2 inch casing reaching down to the Binnewater sandstone, 
depth of 437 feet. The average discharge at about 40 strokes per 
minute was 60-65 gallons per minute, or an average of 90,000 gal- 
lons per day. It will be noted that the discharge was much smaller 
than at hole 32 owing to the absence of gas. Pumping with a 
3/4 inch pump was continued 16 days and 1,532,000 gallons of 



- , _ 


EL 2 SO.' 


2, AO.' 


- I30! 


kv- - - - 
1 906; 

2 2o' 

41 O.' 

— " S^"- 







Fig. 24 Diagram showing successive stages of ground water level between holes 32 and 

34 during pumping 

water were pumped in addition to the 490,000 gallons from the 2^4 
inch pump. The ground water level in hole 34 was lowered 36 feet 
in addition to the 17 feet by the 2 % inch pump, but rose 9 feet in 
20 minutes, and 30.5 feet in the next five days. In the next 22 days 
it rose 9.15 feet, or .42 feet per day. 

Reduced water level in hole 32, 750 feet away by pumping in 34, 
15 feet, or 1 foot for each 120,000 gallons pumped, In the first 



three clays after pumping ceased water rose 5.2 feet, and in 22 days 
rose 9.8 feet or at the rate of 0.45 feet per day. 

During construction 1 shaft 4 located at same point as hole 32/22, 
station 607 + 50, has proved a very wet shaft, the inflow varying 
from 400 to 850 gallons per minute. Pumping at this shaft has 
lowered the general water level and correspondingly lowered the 
water level in hole 34/22 at station 600 + 00. 

Fig. 26 Curve showing rise of water in holes 32 and 34 after pumping ceased in hole 34 

1 From this shaft after reaching tunnel grade, — 250 feet, and after running 
northward into the fault zone and porous shales, the contractors are pumping 
1300 gallons per minute. 



Between the Rondout and Wallkill valleys the aqueduct is to 
follow a tunnel at hydraulic grade which so far as can be seen will 
cut only Shawangunk conglomerate and Hudson River slates. No 
doubt there are many complicated small structures which because 
of the nature of the slates can not be reconstructed. The work 
of tunneling is not advanced far enough to add anything. But 
in the Wallkill valley, where it is necessary again to plan a pressure 
tunnel several hundred feet below grade, a considerable amount 
of exploration has been carried on. 1 

These explorations [see sketch map fig. 8] are distributed along 
several lines crossing the valley at intervals between Springtown, 
about 3 miles north of New Paltz, and Libertyville, which is about 
an equal distance south. 

The geology is simple. Only Hudson River slates form the rock 
floor, and so far as can be judged no other formation is likely to 
be cut by the tunnels. There are no doubt many complicated struc- 
tures, both folds and faults, as indicated by the high dips, but again 
because of the nature of this rock it is impossible to discriminate 
closely enough between different beds to determine exact relations. 
The point of greatest practical importance lies in the fact that 
the rock is fairly uniform and, although much disturbed is of 
such nature that crevices and joints or fault zones are almost as 
impervious as the undisturbed rock. This is because of the tend- 
ency of a formation of this composition to heal itself with fine, 
compact clay gouge. In fact, the mechanical disturbance produces 
or develops the cement filling contemporaneously with the move- 
ment. It is chiefly a mechanical filling, whereas the healing of a 
harder and more brittle rock like a granite or a limestone requires 
more chemical assistance. 

An additional practical question involves the estimate of depth 
required to avoid any possible buried Prepleistocene gorges and 
maintain a safe cover to guard against undue leakage or rupture. 

1 Explorations on the Wallkill division are carried on under the direction 
of Lawrence C. Brink, division engineer. The final construction is in charge 
of James F. Sanborn, division engineer, with headquarters at New 
Paltz, N. Y. 



To this end most of the explorations were made. Two lines less 
than a mile apart on which a few exploratory borings were made 
near Springtown indicate two buried channels, a master channel 
and a tributary from the west which converge northward. A 
maximum depth reaching 70 feet below sea level was found on the 
more northerly line almost directly beneath the present stream 
channel which flows on drift at an elevation of 150 above tide. 

The more southerly profile reaches only sea level indicating a 
gradient for the preglacial stream at this immediate locality of more 
than 79 feet per mile. 

In the vicinity of Libertyville, 5 to 6 miles farther south, where 
the aqueduct was finally located, the profile was found to be con-, 
siderably higher. Intermediate profiles are shown in accompany- 
ing figures. The deepest point yet found on the Libertyville line 
is 65 feet above sea level. It is worth noting that the gradient of 
the ancient Wallkill is therefore shown to be decidedly unsymmet- 
rical. The rock floor formation remains the same although it may 
vary somewhat in character. Under these circumstances, however, 
a gradient of 13 feet per mile from Libertyville to Springtown 
forms a sharp contrast with the 79 feet per mile represented at 
the Springtown locality. In view of the remarkable increase of 
gradient and the narrower form it seems reasonable to regard this 
as a rejuvenation feature developed at the time of extreme con- 
tinental elevation. 

How much deeper the lower Wallkill may be, including the so 
called Rondout river, which is really a continuation of the ancient 
Wallkill and geologically belongs to this drainage line instead of 
to the Rondout, no one can tell. But it is at least interesting to 
observe that the intervening distance from Springtown to the Hud- 
son at Kingston is approximately 12 miles and that a gradient for 
that distance equal to the average known in the 6 miles explored, 
i. e. 24 feet per mile, would depress the outlet 288 feet more. 
That would be equivalent to 367 feet below sea level. If, how- 
ever, a steep gradient such as that at Springtown prevails in this 
lower portion it is necessarily much lower — for example if a 79 
foot gradient is maintained it would be possible to reach a final 
outlet at — 1029 feet. It is likely that an intermediate value is 
more nearly correct. This has, however, an important bearing 
upon the question of maximum Hudson river depth, especially the 
existence of an inner deeper gorge above the Highlands. So far 
as this Wallkill profile goes, it supports the gorge theory. It is 
certain that the Prepleistocene Wallkill flowed north not very dif- 

Plate 25 




,'.■/! L L I L L 5 1PHON 

Cross section showing the buried preglacial Wallkill channel as indi- 
cated by exploratory borings near Springtown 


Profiles of the present and preglacial Wallkill channels near Liberty- 
ville, and a diagrammatic section showing the different types of drift-filling 
together with the borings which furnished the data 


ferently from the present stream except on a steeper gradient, but 
in all probability the headwater supplies between this stream and 
the Moodna have been somewhat shifted. It is possible that some 
former Moodna drainage area is now tributary to the Wallkill. 
But these changes were wholly glacial in origin and the extent of 
such shift is indeterminate at present. 

It is a notable fact that a large proportion of the work of ex- 
ploration in this valley was done successfully by the wash rig. 

The extensive lot of data was gathered without much delay or 
difficulty. This is because of the nature and origin of the drift 
cover. A considerable proportion of the drift mantle especially 
in central and deeper portion of the valley is modified assorted 
sands, gravels and silts or muds. In part they represent deposits 
in standing water laid down at a time when the lower (north) end 
of the valley was obstructed by ice and while waste was poured 
into the valley from neighboring ice fields. It is impossible to 
reconstruct the beds of these materials with any degree of accuracy. 
But it is at least certain that lens or wedgelike layers of differ- 
ent quality of material were penetrated, indicating oscillation and 
overlapping of deposition conditions, boulder beds and till being 
interlocked with assorted sands and gravels. But there is appar- 
ently no evidence of ice deposits of greatly differing age. The 
accompanying profile and cross section is a representation of ma- 
terials on the Libertyville line based upon identifications made by 
the inspector of the Board of Water Supply of the Wallkill Divi- 
sion under Mr L. C. Brink, division engineer. 



Moodna creek enters the Hudson from the west between Corn- 
wall and Newburgh not more than a mile north of the entrance 
to the Highlands. It is a retrograde stream in its backward flow 
similar to the Wallkill. But its channel at present is almost 
wholly on glacial drift which it has trenched to a depth of more 
than 100 feet below the average adjacent surface. How much 
of its retrograde course therefore may be postglacial is not so 
clear. It seems necessary, however, to account for all drainage 
on the north margin of the Highlands by streams flowing to the 
Hudson northward. There is no notch low enough for their escape 
elsewhere. The ancient Moodna must have carried most of this 
run-off from the district occupying the angle between the Wall- 
kill and the Highlands. This stream ma}' have drained even more 
of the region now forming the divide with the Wallkill than does 
the present Moodna. In any case it must have been a stream of 
considerable size, capable of excavating a valley or gorge of 
greater prominence during the period of early Pleistocene rejuvena- 
tion than now appears. Furthermore its position makes it highly 
probable that tributaries of fair size entering in its lower course 
were also effective enough to require consideration. This conclu- 
sion has led to the exploration of the Moodna valley in consider- 
able detail in preparation for the aqueduct work. 

The Catskill aqueduct is to cross the stream near Firth Cliffe, 
which lies almost directly west of Cornwall -an-Hudson, and be- 
cause of the low surface elevation across this valley, as in the 
others, a pressure tunnel in rock is judged to be the most suitable 
type of structure. The accompanying sketch map shows the 

Explorations were conducted especially for the buried channels 
and character of rock floor. 

Geologic features 

The region is one of chiefly Hudson River slate. But there 
are inliers of the older rocks such as Snake hill which belongs to 
a long ridge of Precambric gneiss and granite, brought to the sur- 
face by folding and faulting and there are more rarely outliers of 
younger formations such as Skunnemunk mountain. Farther north 




at Newburgh a gneiss ridge is accompanied by limestone, but in its 
soutberly extension the slates are in direct contact. This relation 
is believed to be wholly due to faulting on both limbs of the anti- 
clines. This gneiss ridge disappears southward beneath the drift, 
but the borings have shown that it continues across the aqueduct 
line, although it has lost its influence on the topography.' There 
are other inliers of similar character such as Cronomer hill 3 miles 
northwest of Newburgh. Between these two gneiss ridges lies the 
southerly extension of the Wappinger limestone belt. But so far 
as is known it disappears beneath the Hudson River series long 
before reaching the line of exploration. 

Near Idlewild station, filling the space between the two branches 
of the Erie Railroad, there is a syncline containing the series of 
Siluric and Devonic strata which spreads southwestward to include 
Skunnemunk mountain, an outlier of Devonic strata. This is the 
only occurrence of these formations in this region south of the 
Rondout valley. The structure and stratigraphic features of this 
occurrence have been worked out by Hartnagel. Its northward 
extension in all probability terminates abruptly by a cross fault not 
far north of the Ontario and Western Railroad. 

From these occurrences southward to tine Highlands proper 
nearly everything to be seen through the drift is Hudson River 

The Highland gneisses are bounded on the north side by a fault 
or series of faults. This brings various members of the overlying 
series into contact along the margin. In the best place where a 
direct observation can be made the gneisses are thrust over upon 
the Hudson River slates along a plane that dips about 40 degrees 
to the northeast. It is probable that a displacement of as much 
as 2000 feet or more could reasonably be assumed at this place. 
The contact zone also is much crushed and bears every evidence 
of having undergone extensive disturbance of this kind. Others 
of this same type occur within the gneisses where weaknesses 
formed in this way permit the development of such notches as 
Pagenstechers gorge. In some cases the rock beneath the surface in 
these zones is more decayed and less substantial than that at the 


The first borings made with the wash rig were found extremely 
unreliable in the Moodna valley. That is because of the very 
heavy bouldery drift forming the greater part of the filling on the 
ancient topography. Next to the Hudson river gorge itself, no 



place has presented greater difficulties in penetrating this drift man- 
tle. Boulders of such immense size occur that they have to be 
drilled like bed rock. In one of the holes a boulder 30 feet 
through was penetrated and 100 feet more of drift found below. 
Progress in such ground is extremely slow and costly. This is 
so much the more so where as in this case there are long stretches 
with unusually deep cover. 

A glance at the accompanying profile and cross section will show 
a very deep and wide valley. Many of the borings are more than 
300 feet in drift which almost wholly obscures the ancient topog- 
raphy. The present Moodna is about half as deep and occupies 
the extreme eastern margin of the older gorge. There is a sec- 
ondary gorge on the west separated from the main channel by a 
sharp divide. A few other smaller notches in the line represent 
smaller tributary or independent stream courses. One of these 
of much interest is known as Pagenstechers gorge. 

The rock floor at all points except two in the central Moodna 
valley including its two nearest tributaries is Hudson River shales, 
slates and sandstones of considerable variation, sometimes much 
brecciated. The two exceptional borings are no. 8/A44 and no. 
16/A44 on the west flank of the westerly tributary gorge, and 
they are in pegmatite and granitic gneiss which is in all probability 
the narrow southerly extension of the Snake hill ridge. Here 
again neither quartzite nor limestone were found on the flank, a 
condition that seems to support the view of a double fault along 
the Snake hill ridge. 

In striking contrast with the broad central Moodna are the two 
narrow and very deep notches farther to the east, the first in 
slates and the second (Pagenstechers) in Highlands gneiss. 

Special features 
Course of the Moodna. The chief interest centers around the 
Moodna channel. There are several unusual conditions, for 
example : 

The rock floor along the profile is almost flat for a distance of 
nearly half a mile in spite of the fact that there would seem to 
be every reason for a different form. The differences in hard- 
ness of rock floor alone would encourage differential erosion ; and, 
since the structure of the formations, the strike, is almost parallel 
to the supposed course of the stream, the influence of different 
beds would be at a maximum. Furthermore, the deep gorge of 
the Hudson, into which the stream flowed is only 2 miles away; 



and if that gorge represents stream erosion to such depth (over 750 
feet) it would indicate a gradient of nearly 300 feet to the mile 
for the last 2 miles of the Moodna — a condition to say the least 
decidedly unfavorable to the development of a flat-bottomed valley. 

Of course, if the profile as determined can be assumed to run 
exactly parallel to the old stream channel for half a mile it would 
be less surprising. But even then it is too flat. For so short a dis- 
tance from the Hudson gorge the gradient ought to be much 
greater than the variation observed in the Moodna channel. There 
are certainly reasons in the structural geology favoring a northeast 
course instead of one parallel to the profile line. And if the 
stream really did flow across this structure, the differences of 
hardness of beds ought to have encouraged a much greater differ- 
ence in depth of channel than the profile presents. With structures 
all running northeast there is every reason to expect the stream 
to follow them. 

Recent exploratory data strongly supports the theory that the 
Hudson gorge at Storm King gap is widened and possibly some- 
what overdeepened by glacial ice. Under normal stream relations 
one might consider the Moodna a tributary hanging valley, itself 
rounded and smoothed to a broad U-shape by ice. This would be 
a very easy solution if it were not for the fact that this tributary 
Moodna opens into the Hudson as a reversed stream, i. e. it opens 
against the flow of the Hudson and more or less directly against 
the known ice movement. It can not be a hanging valley there- 
fore of the normal sort. If a hanging valley of ice origin at all 
it would necessarily be one therefore gouged out by ice moving 
from its mouth toward its head, a case that so far as the writer 
knows has never been observed. The chief objection to this theory 
is that in no other gorge or channel (with one exception, the Hud- 
son at Storm King gap) anywhere in the region so far as known 
is there any evidence of serious modification of an original stream 
channel by the ice invasion. Of course, the axis of the valley is 
favorable and the situation is peculiar in that it parallels the High- 
lands front in this vicinity and the action of the ice may be as- 
sumed to have been somewhat concentrated along this margin be- 
cause of the obstruction. 

Inner notch or secondary gorge. Those who habitually em- 
phasize ice action would no doubt choose to regard this whole val- 
ley as shown in the profile, as chiefly glacial in character and 
origin. Tf that explanation is the true one, then it must be ad- 
mitted that a deeper smaller inner notch or gorge is unnecessary 
and indeed unlikely. 



The critical point therefore in the whole argument is as to the 
origin of the Yalley, i. e. is it essentially a stream valley ? Or is it 
as to present rock floor form wholly a glacial valley ? . 

If it is a stream valley then no doubt full account must be taken 
of the proximity to the Hudson, and the possibility of developing 
a temporary graded condition and some adequate allowance must 
be made for its work during the subsequent continental elevation 
and the deepening of that river to several hundred feet below the 
known bottom of the Moodna. In short, one would expect a nar- 
row deeper notch in the Moodna floor as a result of this rejuvena- 
tion. But on the contrary if in preglacial time the stream were 
not so powerful and had not been able to keep pace, and if the 
ice movement can be assumed to have concentrated along this line 
to such efficiency as to gouge out a groove 3000 feet wide almost 
flat to a depth of 300 feet only guided in direction by the original 
Moodna, then one may readily abandon the idea of a deeper notch. 

One or the other of these types of origin must be the chief 
factor in reaching a reasonable opinion as to the presence of an 
inner notch. 

In any attempt to choose between these factors, one is led -to 
reconstruct the preglacial drainage lines. When this is done it at 
once appears as most probable that there was at that time as now 
a considerable area tributary to the Hudson with a stream course 
very much like the present Moodna. In other words a fair sized 
stream is assured. Once such a stream is granted and the effects 
of its work reckoned in full knowledge of the adjacent Hudson, 
and its probable behavior is studied in the light of the data ob- 
tained in exploration of the valleys of other tributaries, it becomes 
more and more difficult to wholly eliminate the inner gorge idea. 
It seems to the writer probable that the valley owes its erosion 
chiefly to the preglacial stream. But the channel has suffered sub- 
sequent widening and smoothing by ice especially in its upper and 
broader portion, below which there may yet be a notch. One must 
admit that the results of boring prove the notch to he very nar- 
row, less than 150 feet, or else not there at all. In reaching an 
opinion as to the possibility of one so narrow, it is worth while 
to note that the Esopus, which is a larger stream, has cut down 
at Cathedral gorge to a depth of from 50 to 80 feet with almost 
vertical sides and only about 150 feet wide. This gorge further- 
more is cut in almost horizontal strata of such character that 
there is no special structural tendency in them to contract the 
stream. At the Moodna on the contrary, in addition to the smaller 



size of stream, the rocks stand on edge and run parallel to the 
supposed course so that this structural influence is toward a nar- 
row and reasonably straight gorgelike form. It is not only pos- 
sible that the gorge is narrow, but even probable that it is narrower 
than the present Moodna, i. e. less than 100 feet wide. 

How deep such an inner gorge may be if it does exist is a prac- 
tical question in this particular case, because its depth has a direct 
influence on choice of depth of pressure tunnel. Because of the 
evident narrowness it is likely that it is not of very great depth 
— in view of the quality of these shales perhaps not over a hun- 
dred feet. 

Is there any one point more than another favorable for such a 
notch? There are two facts bearing on this question, (i) the vari- 
ation in core saving which indicates that hole no. 5/A44 with 
7<t has a recovery of only 1/5 the average, and (2) the fact that 
hole no. 15/A44-I- , which is the next hole, shows the lowest bed 
rock in this valley. On the ground of profile therefore and on the 
ground of structural weakness there is reason to choose this space 
between no. 5/A44 and no. 15/A44 as the most likely position. 

Summary. The very abnormal profile of the Moodna valley 
based upon the borings may be due either (1) to parallelism with 
the stream course, or (2) to a graded condition of the stream in 
some preglacial epoch, or (3) to modification of an original less 
prominent channel by ice erosion. 

It is the opinion of the writer that the ancient stream crossed the 
profile line much as the present stream does, that the additional 
narrower valley immediately to the west side is that of a pre- 
glacial tributary instead of a bend of the Moodna itself, that there 
was a development of a moderate sized somewhat flattened valley 
corresponding to the benches and shelves noted in other streams, 
including the Hudson, that subsequent elevation of the continent 
rejuvenated the stream which cut a deeper narrow inner notch, that 
glacial ice moving in reverse direction widened and smoothed this 
upper portion of the valley, but that there may yet be a remnant 
ot the deeper notch in its bottom, and that the space between holes 
no. 5/A44 and no. 15/A44 is the most likely location of this inner 
gc rge. 

Tributary divide. The sharp divide between the two deep 
portions of the valley bottom has proven an evasive feature in the 
later exploration. Two holes put down a short distance to the 
southward (24/A44 and 20/A44) failed to find the rock floor so 
high, one reaching rock at a depth of 181 feet and the other failing 



to find rock even at 213 feet. Two others nearly a thousand feet 
to the westward, however, found rock again at approximately the 
same elevation as the divide. If this is a tributary stream divide 
therefore it must have an east-west trend. 

Pagenstechers gorge 

This is a notch between Storm King ridge and Little Round top 
occupied by a very small mountain stream. The rock floor is granite 
gneiss of the Storm King type. Its special characters are (1) 
extreme shattering or crushed condition, and (2) extensive decay 
along this zone which has softened the rock constituents to great 

Considering the nature of the granite gneiss in general this nar- 
row gorge is a surprisingly deep one. But this is no doubt due to 
the influence of the decayed crush zone. The drill cores taken from 
the holes that penetrated the floor at this place are so much altered 
that, after several months exposure to the air, they can be readily 
crushed in the hand. Hole no. 16/A45 which is centrally located 
penetrated to — 196 feet. It is in material of this same condition, 
to at least — 100 feet. Similar conditions are proven to the north 
of the line, shown in the accompanying profile and a rapid increase 
in depths. From the surface outcrops farther up the gulch it is easy 
to see that the crushed zone extends in that direction with the 
strongest lines about s. 70 w. This is doubtless on the strike of the 
fault lines of the northern border of the range. It is of more than 
usual interest in showing the depth to which incipient decay has 
penetrated in these crush zones, and the efficiency of stream erosion 
along them. 

Overthrust fault 

The principal fault line follows the margin of the granite gneisses. 
At the best exposure of it the Hudson River slates are overridden 
by the gneiss. This represents therefore the cutting out entirely of 
the Wappinger limestone and the Poughquag quartzite and a part 
of the slates by the displacement which must amount to at least 
2000 feet and probably more. The same relation is indicated by the 
borings and by the outcrop near the village of Cornwall, but a little 
limestone is found midway between the two points along the strike 
of the fault. The strike of the fault averages about n. 65 to 70° e., 
but locally, at the best exposure, it is only n. 35 e. The dip is 
southeast at an angle of approximately 45 degrees. 




Moodna valley 



elevation in 

elevation in 

in feet 


cent core 

Kind op rock 



+ 27 


Slate and sandstone 

(On porosity test with plug at 58 feet deep the loss of water was 6 gallons 
per minute with pressure of 0-10 pounds per square inch.) Test unsatis- 
factory because of large hole. 

2 I 236.5 I ? o I o 1 o 

' 3 1 36-3 ' 3 6 - 7 ' 26 I o I Slate 

(On porosity test with depth to plug 173.5 feet and pressure 0-60 pounds 
per square inch the loss was 5 gallons per minute.) Test unsatisfactory 
because of large hole. 









Slate and sandstone 









Slate and sandstone 







+ 26 















cent core 

Kind of rock 



in feet 

saved 1 




















'~ 4 








Slate and sandstone 

■:■ 5 




















1 10 



Slate and sandstone 





2 • 



Slate and sandstone 


1 76 













1 1 

169 . 






1 2 

221 . 






Slate and sandstone 


226 . 







226 . 




3 1 • 

1 5 

230 . 




3 2 - 


Slate and sandstone 






32 • 



169 . 





1 In cases which show no recovery of core a method of drilling was employed different 
from the others and the rock was ground to pieces. Failure to recover core may therefore 
be no indication of poor rock quality. 













cent core 

Kind of rock 



in feet 


1 l 



— r 7 





Slate and sandstone 

l 2 

2 74 


+ 171 

• 5 




Slate and sandstone 




+ 39 









— 22 





Slate and sandstone 




— 47 

5° • 



Slate and sandstone 




— 40 






299 . 


— 5i 


102 . 






+ 09 

109 . 



Granite gneiss and 


282 . 


+ 7 



J 3 

Slate and sandstone 








Slate and sandstone 

1 1 







I I 

Slate and sandstone 


272 . 





Slate and sandstone 







1 3 







91 . 


3 2 

Slate and sandstone 


301 • 






Slate and sandstone 




+ 25- 


104 . 



Pegmatitic granite 


300 . 

— 42 ■ 




Slate and sandstone 

1 Porosity test made on hole no. 1 shows a loss of .03 gallons of water uider 100 pounds 
pressure with packer at depth of 3S7 feet. Depth to grjun 1 water 217 feet. 

Porosity test on hole 2/A44. 

Ground water level at a depth of 90 feet = el. + 184.5'. 


Depth to packing 
in feet 

1 46 
1 96 









1 00 

1 20 

• 2S 




■ 79 

. 20 


• 3S 




. 1 2 


• 19 


100 = Gage pressure 

140 = Calculated pressure 1 

1 .03 = gallons lost 


'Calculated pressure equals average pressure plus weight of column of water from surface 
to ground water level. Gage pressure is given in pounds per square inch. Loss is in gallo n s 
per minute. 












cent core 

Kind of rock 



in feet 







1 9 



Slate with quartz 












2 S 



43 2 




















Slate with quartz 









2 14 






Decayed granite gneiss 















1 1 

2 I 4 






Decayed granite gneiss 





























Decayed granite gneiss 




1 1 





Decayed granite gneiss 
and seamy gneiss 








Gneiss and dyke rock 



Foundry brook is a small stream entering the Hudson at Cold 
Spring in the Highlands. It drains a rather abnormally large valley 
bordering Bull mountain, and Breakneck ridge on the east, and its 
axis is in the strike of the principal structure of the gneisses which 
form the chief rock formation of the floor. This valley is in exact 
line with the course of the Hudson from West Point immediately 
southward, and its rock formations are similar in character and con- 

There is greater variety of rock composition in this belt, i. e. the 
Foundry Brook-Hudson river belt, than in any other in the High- 
lands of similar area. The eastern half of the belt is a typical 
development of banded gneisses and schists and quartzites belonging 
to the sedimentary representatives of the Highlands gneiss. Small 
layers of interbedded limestones are frequent together with serpen- 
tine, and mica and graphite and quartz schists. In addition along 
the east bank of the Hudson, they are profoundly modified by 
crushing and shearing in zones that trend with the formation, i. e. 
in a direction leading toward and through Foundry brook valley. 

The west side is much less variable and is bounded at the margin 
by one of the most massive types of the region — the Bull mountain 
and Breakneck mountain gneissoid granites, which are essentially 
the same as that of Storm King mountain. 

But additional structures enter Foundry brook valley from the 
western side at an acute angle with its axis and formational trend. 
These additional structures are two well marked faults, which cross 
the Hudson — one along the precipitous southeast face of Crows 
Nest and the other along the southeast face of Storm King moun- 
tain. These are the most pronounced escarpments of the whole 
region. The first one crosses the Hudson between Cold Spring and 
Bull mountain and in passing northeastward loses much of its in- 
fluence upon topography and its movement is probably dissipated in 
that direction. A line from the southeastern face of Crows Nest to 
the point to be described runs n. 50 e. 

1 Explorations at Foundry brook were clone under the direction of Mr 
William E. Swift, division engineer, now in charge of the Hudson River 
division of the Northern aqueduct. 

6 163 




Foundry brook therefore contains structures that could produce 
considerable effect upon the quality and condition of rock floor. 
The rock floor is covered with heavy bouldery drift — thicker on the 
Bull mountain flank than in the valley bottom proper. Where the 
aqueduct line crosses the floor is at an elevation of 200 feet to 350 
feet A. T. Hydraulic grade of the aqueduct is about 400 feet. 

The lowest bed rock found along the line is 182.3 feet and the 
channel of the present stream coincides with the preglacial one in 
that portion of its course. There are two secondary channels — 
probably tributary stream channels on the west side. One of these 
lies under 70-80 feet of drift. 

Borings were made for the purpose of determining the rock floor 
profile and the condition of bed rock. In most of them the ordinary 
gneisses and granites were penetrated in normal condition. 

But in a few a very unusual condition was found. Hole no. 2 at 
el. 347 feet near the west or Bull mountain margin penetrated 49 
feet of drift to el. 298. Then the drill passed into gneiss which was 
at the top, the first 30 feet, of a fair quality. This is shown by the 
core recovered — the first 12 feet -recovering over 50^. But the 
percentage of recovery rapidly fell off — amounting to only 36$ in 
the first 50 feet. Only 1 foot of core was recovered in the next 30 
feet, or only 3^. While from that point el. 220 feet to the bottom 
of the hole el. 51.8, at a depth of 295.7 feet from the surface, 
nothing but fine decomposed matter was washed up. There was no 
core at all. This was at first reported as sand by the drillmen, and, 
coming at a time when exploration of deep buried gorges was the 
rule at other points of the aqueduct, there were many questions 
about the interpretation of this new hole, the first assumption of 
the drillers being that an overhanging ledge of a very deep gorge 
had been penetrated passing through it into river sands below. A 
little study of the material proved that this view is untenable. The 
sandy wash from the drill is true disintegrated gneiss much decayed 
and dislodged by the drill. 

But the meaning of it and the extent of it are after all important 
additional questions. 

Interpretation and further explorations 

It is certain that the soft material and the " sand " reported from 
this boring represent rock decay induced by underground water 
circulation. Water circulation is rather free as is shown by the 


fact that there was an artesian flow from this hole of 10 gallons per 
minute after reaching a depth of 80 feet, which increased to 15 
gallons per minute after reaching a depth of 253 feet. This under- 
ground supply is maintained since completion and the pressure is 
sufficient to raise the water about 15 feet above the surface. 

This is a behavior that is consistent with the geologic conditions. 
The boring has no doubt penetrated a crush zone following one of 
the faults which enters this side of the valley. The crush zone dips 
steeply and the boring has penetrated the hanging wall of more 
solid rock in the first 50 feet and, after reaching the broken and 
decayed portion of the zone, has swung off parallel to the dip and 
avoiding the more resistant foot wall has followed down on the soft 
inner streak. 


Fi° 30 Sketch illustrating the interpretation of geologic structure across Foundry 
brook valley indicating the relation of certain borings to them and their supposed 
influence in deflecting the drills ^^ttiimt^J 1 

This crush zone extends on northeastward across higher ground 
where opportunity for taking in surface water is offered. This is 
without doubt the source of supply for the circulation which fur- 
nishes the artesian flow and which has been so effective in pro- 
ducing decay to great depth. But the circulation and associated 
decay are probably limited to comparatively narrow zones. There 
is no good reason for assuming large masses of rotten gneiss at 
great depth. The worst zones are narrow but may be comparatively 
deep, i. e. they may extend much deeper than any of the borings yet 
made in this valley. The depth of decay is related to the outlet for 
underground circulation which in this case is the gorge of the 



Several other boring's encountered similar conditions, especially 
those on the west flank of the valley within range of the belt in 
which the fault seems to be located. 

Hole no. 9 reached the rock floor at a depth of 80 feet, and then 
penetrated rock to a depth of 159.7 feet. All of the material is 
badly decayed. Only 1 foot of core was recovered from the whole 
boring and that is mostly quartz coming from a veinlet or peg- 
matitic streak at 141 feet. Water under slight pressure was en- 
countered in this hole also. But because of the somewhat greater 
elevation of the surface at this than at hole no. 2 there is not a 
constant outflow. 

Two other holes immediately to the west show much better rock 
condition — no. 1 showing 79$ core recovery. Also two on the east 
side at greater distance [see accompanying profile] show good rock. 
But one other no. 3 at a distance of over a thousand feet to the east 
encountered another zone of decayed rock, the record being very 
similar to no. 2 in that poorer conditions are shown at depth than 
near the surface. Rock was found at a depth of 20.2 feet. From 
20.2 to 116 feet the gneiss was quite hard, 55.3 feet of core being 
recovered or S7-7/'- But from 116 feet to the bottom 207.5 feet the 
material was as bad as in hole no. 2, and no core was recovered. 

Several other tests were made on the borings with a view to de- 
termining the character and extent of these features more definitely. 
For example, if the interpretation given for the behavior of no. 2 
and no. 3 is correct it ought to be possible to survey the holes and 
determine a deflection from the vertical as the drill deviated from 
its course to follow the softest streak. A survey conducted for this 
purpose indicates just such a result. The accompanying sketch 
shows the data plotted. The drill was deflected 4 36' at a depth 
of 50 feet, 7 36' at 100 feet, 8° 2' at 150 feet and 9 40' at 
198 feet. 

Pressure tests were made for porosity on some of the holes in 
sound rock. Some of these data are given on the profile. 

Some of the rock of this valley, if very extensive, such as that 
in borings no. 2, no. 3 and no. 9, would be very poor ground for 
tunneling. The practical question involves especially the width of 
these zones, are they a foot wide or are they a hundred ? In an 
attempt to help settle that question an inclined hole was proposed 
that was to run at an angle low enough to crosscut these belts. 
Accordingly hole no. T4 was bored inclined 40 26' to the hori- 
zontal and started on the solid gcanite gneiss. The results were not 



Figure 31 



very encouraging. The decay is shown not to be confined to mere 
seams. The doubt raised by so much bad ground has finally led to 
the adoption of a different plan for crossing Foundry brook valley 
and no further data are likely to be added by this work. As it now 
stands the borings at Foundry brook indicate the deepest decay of 
any yet made in granites or gneisses except those of Pagenstechers 
gorge on the north side of Storm King mountain. Both cases are 
of similar origin and history, but Foundry brook is apparently the 
more complex in occurrence. There are several parallel zones 
along which there is extensive decay to a depth of more than 300 



Three creeks unite to form an inlet at the sharp bend in the 
Hudson immediately above Peekskill. The middle one of these 
is known as Sprout brook. It occupies a deep and narrow valley 
that is well marked for 10 miles in its lower course and is trace- 
able as a physiographic feature of less prominence to the north mar- 
gin of the Highlands. Its persistence indicates some important 
structural control in erosion. 


This valley lies in the midst of the most typical gneisses and 
granites of the Highlands region. And in addition several of the 
" iron mines " of Putnam county lie on its western flank. The 
rocks are complex granitic and quartzose gneisses and granites. 
Foliation and banding and bedding wherever this appears is parallel 
to the axis of the valley. The most notable geologic feature is the 
occurrence of a broad belt of crystalline limestone throughout the 
lower 4 miles. It is undoubtedly chiefly this limestone, which is less 
resistant to weather than the gneisses, that controls the form and 
size of the valley. As to geologic relations, this is one of the most 
interesting formations of the region. It is coarsely crystalline, full 
of silicious impurities at many places and carries small igneous in- 
jections and dykes, and so far as the bedding can be followed, 
stands almost on edge. Although an actual contact is not seen, at 
several places the limestone and gneiss approach within a few feet 
of each other and it is certain that no other formation can come 
between them. This is more certainly indicated in the northerly 
extension of the valley where the limestone gradually disappears 
leaving only the gneisses and granites. That there may be a fault 
contact must be admitted, but of this there is no good evidence in 
the field. 

Such relations and character show that this limestone is similar 
to the smaller interbedded occurrences noted frequently with the 
gneisses in the Highlands and elsewhere. If it is of that type then 
it is the largest representative yet found in that series. But it is 
also in these characters similar to the Inwood limestone of more 
southerly areas. The overlying Manhattan schist which is lacking 




may have been removed in erosion. One of these types it resembles, 
but it can not be the Wappinger (Cambro-Ordovicic) as was 
pointed out by the writer in a former report. 1 The Wappinger, 
wherever known to be such, is never intruded and always lies above 
a thick quartzite (Poughquag). It does so even in the next valley 
(Peekskill creek) less than a mile distant. With the interpretation 
of this Sprout Brook limestone therefore is involved the correlation 
and interpretation of the age of much greater areas. That the 
Sprout Brook limestone is not Wappinger is clear enough, but it 
could be either interbedded (Grenville) or Inwood. If it is Gren- 
ville then of course it has no direct bearing on the Wappinger- 
Inwood question and these two might be equivalents. But if the 
Sprout Brook limestone is not Grenville (interbedded) then it must 
be Inwood and in that case the Inwood and Wappinger are not 
equivalent — which means that there are two series above the 
gneisses instead of one — an Inwood-Manhattan series, and a 
Poughquag- Wappinger-Hudson River series. At the present time 
it is not possible to give with certainty a final interpretation of the 
Sprout Brook limestone. 

Explorations 2 

It was at first believed that a pressure tunnel could be con- 
structed advantageously at the point of crossing this valley and 
borings were made to test rock conditions. The data gathered in 
exploration are indicated on the accompanying geologic cross sec- 
tion [fig. 32]. 

Borings indicate that the rock floor has been eroded to a few 
feet below present sea level and that the gorge has a drift filling 
of more than 150 feet. The central borings penetrate limestone 
and indicate a total width of this type of more than 400 and less 
than 600 feet. The best estimate on the basis of these explora- 
tions is 500 feet. Whether this width represents one thickness 
of the formation as would probably be the case if it is an inter- 
bedded Grenville layer, or part of a double thickness due to infold- 
ing, as would probably be the case if it is the Inwood, there is 
no evidence. The thickness seems to be even greater farther south 
in the same valley (it becomes )A mile wide), but it can not be 

1 Structural and Stratigraphic Features of the Basal Gneisses of the 
Highlands. N. Y. State Mus. Bui. 107 (1907). p. 361-78. 

2 Explorations at Sprout brook are in charge of Mr. A. A. Sproul, division 
engineer in charge of the Peekskill division. 





accurately measured and there is no way of guarding against repe- 
tition of folds. The valley floor is decidedly terraced at an ele- 
vation of about 130 A.T. One side is limestone and the other is 
granitic rock. This is probably a local mark of the Tertiary base 
leveling work. 

Because of the great depth of this narrow gorge, it would require 
a 500 foot shaft at each side to lead from hydraulic grade down to 
a safe level for the pressure tunnel. For a crossing not more than 
2000 feet long this is excessive and the cost becomes greater than 
by other methods of construction. Consequently the tunnel plan 
has been abandoned and it is not likely that further data bearing 
upon these questions will be added. 



Immediately east of Sprout brook, described in the previous sec- 
tion, is Peekskill creek, which drains the largest valley emerging 
from the southern margin of the Highlands. This valley as a 
physiographic feature is continuous with the Hudson river gorge 
from the sharp bend at Peekskill to Tompkins Cove. There are 
important structural features along the strike of this valley which 
extend very far beyond the limits of Peekskill creek itself, among 
which are strong folding and block faulting. The chief fault con- 
tinues to the southwest with still greater prominence and appears 
on the west side of the Hudson in the escarpment forming the 
southeastern margin of the Highlands continuously for many miles 
into New Jersey. 

Near the Hudson, Peekskill creek and Sprout brook unite and 
the structures and formations characteristic of each valley converge 
until in the last half mile of their united course rock formations 
characteristic of Sprout brook lie on one side of the valley, those 
characteristic of Peekskill creek on the other, and the contact which 
follows the divide to that point then passes beneath the waters of 
Peekskill inlet. Because of the block faulting which has carried 
down overlying formations and protected them from the total de- 
struction characteristic of the central Highlands region this valley 
is of unusual interest. 

Explorations 1 

The aqueduct line crosses this valley about 3 miles from the 
Hudson, and in determining the possibility of crossing by pressure 
tunnel in rock a considerable number of explorations were made. 

Enough has been done to outline the rock floor profile very defi- 
nitely and to determine the condition of the formations. 

An examination of the drill cores and records of explorations 
shows the following facts which are compiled as fully as possible 
on the accompanying cross section. 

Phyllite. One boring (no. 1) is in a phyllite whose character 
and relation to other formations leads to the conclusion that it 

1 These explorations were directed by Mr A. A. Sproul, division engineer 
of the Peekskill division with headquarters at Peekskill, N. Y. 


r 7 6 


belongs to the Hudson River slate series. This type of rock forms 
the whole western side of the valley for several miles. Beds stand 
on edge or dip steeply southeastward and are in good sound physi- 
cal condition. The rock is everywhere a fine grained micaceous 
slate or phyllite and in some places carries pyrite crystals. It is 
impossible to estimate the thickness or minor structural habits. 
But it is clear that it forms the upper member of a series that 
has a synclinal structure and therefore the belt represented by 
the phyllite marks the axis of the syncline although the chief val- 
ley development lies wholly to one side. 

Limestone. Eleven borings (no. 2, 3D, 4 C, 11, 13 C. 18, 
22, 23, 25, 26 and 29; are in limestone. All show essentially a 
very fine grained close textured crystalline gray or white or bluish 
rock with strong bedding standing nearly vertical or at very high 
angles. This, because of its character and relation to other forma- 
tions, is regarded as the Wappinger limestone — a formation well 
known north of the Highlands, where it is at least 1000 feet thick. 
From present explorations it is now certain that a belt 3250 feet 
wide is underlain continuously by this formation standing nearly 
on edge. Unless repeated of course this would represent approxi- 
mately the thickness for Peekskill valley. But it is not believed 
to be so thick. It is more likely that there is a threefold occur- 
rence brought about by close isoclinal folding (a closed s-fold) 
as seen in the accompanying cross section. This view is supported 
by at least one occurrence of the underlying quartzite member near 
the center of the valley at a point a couple of miles farther north 
On the line of exploration, however, none of the borings pene- 
trate any other formation beneath. Attention is called to additional 
structural details and physical conditions in a later paragraph. 

Quartzite. One boring (no. 5) is in a quartzite. It i-~ very 
pure, fine grained, closely bound and typical quartzite. The beds 
stand almost vertical aud the whole thickness is known from nearby 
outcrops to be approximately 600 feet. From its character and re- 
lations to other formations it is regarded as the Poughquag — a 
well known formation of the north side of the Flighlands. 

Gneisses. Five borings (no. 7 K, 9 B, 17, 27 and 28) are in 
gneisses. These are to a considerable extent simple granite gneisses 
of igneous origin. But there is the usual variety characteristic of 
the Highlands gneisses and no doubt they are representatives of 
the great basal gneiss series that is elsewhere referred to as the 
equivalent of the Fordham of New York city. 



2 Stratigraphy 

This is therefore the rock series of Peekskill creek. It is 
the only locality on the south side of the Highlands where all 
are represented in complete and simple form. There is no doubt 
that it is the Poughquag-Wappinger-IIudson River series, in spite 
of the complete absence of organic evidence. A similar though 
not so complete and clear occurrence is to be found on the west 
side of the Hudson near Stony Point and Tompkins Cove. That 
is a part of the same structural syncline. It is probable also that 
the phyllite so finely developed in the village of Peekskill in the 
next small valley to the east is the same. But outside of these 
occurrences there are none that clearly represent this same series 
as a whole and in the same condition. 

No more striking example of this fact can be found than the 
adjacent Sprout brook described in an earlier section. There coarse 
crystalline and injected and impure limestone occurs alone — no 
phyllite and no quartzite. When one remembers that the two 
occurrences so strongly contrasted. Sprout brook and Peekskill 
creek, converge until they actually unite, and still preserve their 
stratigraphic dissimilarity, without any adequate structural reason 
for it, the only conclusion possible is that the two occurrences rep- 
resent two entirely different series of formations. 

The Peekskill valley series is Cambro-Ordovicic in age ; what is 
the other? It is older, at least that is certain. But is it (the Sprout 
Brook limestone) as old as the oldest of the gneisses themselves 
and therefore interbedded with them representing locally the Gren- 
ville ; or is it intermediate — Postgrenville and Precambric — with 
which possibly other occurrences of rocks of similar habit and 
equally uncertain relations belong? 

It is on the general similarity of this occurrence to the Inwood 
limestone as known throughout Westchester comity and New York 
city that a tentative intermediate series has been recognized. This 
is the Inwood-Manhattan series. Whether it is in reality a separate 
older series is not regarded as proven. But for engineering and 
practical purposes the distinction is a good one and eminently ser- 
viceable. Further, discussion may better be continued in a different 

3 Rock surface 

The bed rock surface is pretty well outlined by the borings. A 
profile based upon them seems to leave no unexplored space of suf- 
ficient extent to admit a gorge of great consequence to a lower level 

i 7 8 


than is already shown in holes no. I and no. n [see profile and 
cross section, fig. 33]. The elevation indicated by no. 3 D is be- 
lieved to be misleading because of the use of a drill that was 
capable of destroying a part of the ledge rock that would usually 
core. The points believed to be weakened by structural disturbance 
and therefore most likely to be attended by erosion and stream 
action are in the vicinity of hole no. 11, near the present creek, and 
hole no. 25, near Peekskill Hollow road. 

4 Buried channels 

From the accompanying cross section it will be seen that the 
drift cover is more than 100 feet thick over large portions of Peeks- 
•kill valley. The rock floor in the middle of the valley averages 
approximately 25 feet A.T., while the drift surface except where 
cut out by stream erosion is at about 125 feet. In the rock floor 
there are two depressions, the large one wholly within the lime- 
stone belt and the smaller following the limestone-phyllite contact. 
There is not much difference in their depth so far as explored, but 
there is a possibility of a somewhat deeper notch in each one. The 
depth to which some of the limestone beds are decayed by under- 
ground circulation would lead to the belief that a deeper notch may 

The drift cover is chiefly partially assorted sands and gravels in 
the central portion of the valley, and more of a till on the eastern 
valley side. It is noteworthy that the present Peekskill creek lies 
far to one side following closely the phyllite wall. 

5 Underground water 

Present elevation above sea level is so slight that there is appar- 
ently little encouragement of deep underground circulation. But 
ar certain points the rock has been found to be very badly decayed 
to a great depth — to at least 200 feet below sea level. This is 
believed to have been accomplished chiefly at a time when the re- 
gion stood at a higher level. Hole no. 22 is especially notable in 
this connection. A comparison of the figures of core saving is one 
of the best lines of evidence on this question. Wherever data are 
at hand the percentages of saving have been put on the cross sec- 
tion. Hole no. 29, for example, shows a saving of only n<£ in the 
lower 250 feet, reaching a depth of 297 feet below sea level. 

The present water table profile is shown on the cross section. A 
great body of water stands in the assorted sands directly upon bed 



rock forming essentially a great reservoir of supply that has ready 
access to the almost vertical limestone beds. This will give a maxi- 
mum water supply to holes that penetrate porous or broken por- 
tions of bed rock. The attitude of all strata is especially favorable 
for admitting an almost inexhaustible supply from a considerable 
drift-covered area within which circulation is probably very rapid. 

• 6 Condition of rock 

All strata of this valley stand so nearly on edge that drills actually 
explore a very limited portion of the whole series of beds. No very 
great advantage is gained by excessively deep boring because the 
drill follows necessarily almost the same bed from top to bottom. 
At best only the immediately adjacent beds are penetrated. This 
means that much of the total thickness of beds is untouched by 
present explorations, and must be interpreted on the basis of their 
general likeness to those more fully determined. The usual suc- 
cession of beds is known to be quite uniform in quality and loca- 
tions where they can be studied and there is no reason to expect 
greater variation here. 

Deviations from such normal or uniform conditions are mostly 
due (a) to local development of mica from recrystallization of im- 
purities in the limestone, and (b) to crush zones developed in the 
process of folding and faulting which has broken the rock or weak- 
ened it enough to permit a more ready circulation of underground 
water. Wherever either of these structural conditions prevail, the 
rock has been excessively decayed, or disintegrated, or sufficiently 
weakened in its binding matter or its sutures to crumble in the 
hand or break down to a sand under ordinary boring manipulation. 
This condition is known to reach to -297 feet. How much deeper 
is not known. Probably the decay dates back in large part to pre- 
glacial continental elevation at which time probably there was more 
ready deep circulation with possible outlet in the Hudson gorge. 
This action has been all the more effective by reason of the attitude 
of the beds. They stand so nearly on edge that they present all 
their weaknesses of bedding and sedimentation structures to the 
destructive surface agents. They admit surface water readily and 
favor abundant underground circulation. 

Considerable faulting occurs. The contact between the granite- 
gneiss and quartzite is a fault contact. Wherever seen this is sound. 
But a crush zone in limestone lies nearly central in the valley, cut 
by holes no. 23 and no. 25, where the rock shows a finely brecciated 




condition some portions of the drill cores being literally crushed to 

In one hole, no. ri, near the phyilite-limestone contact, a soft, 
sandy condition was encountered at a depth of 133 feet, permitting 
the drill rods to be pushed down without boring at all, 60 feet 
ahead of the casing. This, however, is not believed to indicate any 
very extensive weakness. It is probably connected with the bedding 
planes or joints rather than with general decay or faulting. Four 
or five inches of solution and disintegration along bedding planes 
would account for all that has been proven. The fact that the rods 
could be shoved down 60 feet while the corresponding outer casing 
could be shoved down only half as far seems to support this view. 


If a tunnel were made across this valley there would be approxi- 
mately 1 100 feet of it in Hudson River slate (phyllitc), 3250 feet 
in Wappinger limestone. 600 feet in Poughquag quartzite, and the 
rest in the gneisses. 

Some weak rock is certain to be found, especially in the vicinity 
of station 367+50 and 345+00 to 350-00. At both places increased 
water inflow would be encountered with almost exhaustless supply 
from the sands that lie on the rock floor above. 

At about this stage in the exploration the Board of Water Supply 
decided to abandon the rock tunnel plan. The conditions found 
were considered by them too questionable. Steel pipe construction 
is to be substituted. As a result it is not likely that much more 
detail will be added to the structure of this very complex valley. 



It is proposed to finish Ashokan reservoir and the Northern aque- 
duct first. This so called Northern aqueduct reaches from the Cats- 
kills to Croton lake. Croton lake is the present supply of New 
York city and is already connected by two aqueducts with the city 
distribution. As a first step, therefore, and as an emergency meas- 
ure the Catskill water will be delivered to the Croton system by 
finishing the Northern aqueduct first. As rapidly, however, as the 
whole project can be carried out the so called Southern aqueduct 
will be constructed to continue the Catskill water independently of 
the Croton supply to the city. 

The Southern aqueduct department has charge of the line from 
Hunters brook on the north side of Croton lake to Hill View reser- 
voir near the New York city boundary. During exploratory work 
it has been under the direction of Major Merritt H. Smith, depart- 
ment engineer, with headquarters at White Plains. Construction 
now going on is in charge of Mr F. E. Winsor, department engineer. 

The first link in this southerly extension is to be a tunnel be- 
neath Croton lake through which the Catskill water may pass in the 
same manner as it crosses other valleys. This crossing has been 
located a short distance below the old dam on the Croton, about 5 
miles up stream from the Hudson. 

The problems involved at this point include ( 1 ) a determination 
of the kinds and quality of rock to be penetrated, (2) their water- 
carrying capacity, and (3) opinion as to the proper depth for a 
successful tunnel. 

Geological features 

The Croton valley is one of the very few in southeastern New 
York that actually crosses the geological formations and major 
structural features instead of following parallel to them. In its 
lower portion it passes from gneiss to limestone and to schist sev- 
eral times. The reason for this somewhat abnormal course is prob- 
ably the development of weak zones by fault movements in this 
transverse direction. 

Only one of the well known formations of rock is exposed in 
the immediate vicinity of the tunnel site. This is the Manhattan 
schist, the uppermost formation of the region south of the High- 
lands. Along the Croton it varies greatly, the chief type being a 



l8 5 

garnet-bearing quartz-mica sohist varying from rather fine grain 
and semigrannlar appearance to a very coarse and strongly foliated 
structure. This part of the formation undoubtedly represents re- 
crystallized or metamorphosed sediments. But associated with this 
fades there is a more dense black hornblende schist that, not only 
here but at many other places, is thought to represent igneous in- 
trusions that have been metamorphosed together with sediments of 
various types, until both have lost almost all of their original char- 
acters. The hornblendic schist type is not so extensive as the other, 
the mica schist, but it is more compact and here as usual is in the 
better condition. 

Pegmatite stringers occur abundantly, especially in the mica schist 
varieties. They are of no great consequence, however, as a factor 
in this study. They originate! in the aqueo-igneous activity in- 
volved in the reerystallization of the rock when it was worked over 
into a schist. 

Beneath this Manhattan schist formation lies the Inzvood lime- 
stone, a bed probably at least 70c feet thick. But at this point how 
deep it lies and at what depth it would be penetrated nobody can 
tell. None of the drills have touched it. Beneath the limestone in 
turn lies the granitic and banded gneisses belonging to the Fordham 
gneiss series, the lowest and oldest of the region. 

Along the Croton river nothing but Manhattan schist is to be seen 
at the surface for more than a mile above and below the proposed 
crossing. The same thing is true for an equal distance on opposite 
sides from the river at this locality. 

But the structure is folded and the normal northeast-southwest 
trend of the folds crosses the river, every arch or anticline tending 
to bring the limestone and gneiss nearer to the surface. One of 
these folds does expose the limestone and gneiss in a strip extend- 
ing from the Hudson river northeastward for two thirds of the 
distance to the old Croton dam. But before reaching the Croton 
valley this fold pitches down toward the northeast beneath the Man- 
hattan schist and passes under the present lake (or reservoir) in 
diat relation, nut reaching the Mirface again Eor a distance of about 
6 miles. At least one more fold is known to behave in a similar 
manner as it reaches the Croton. 

These facts make il certain that there is limestone beneath the 
schist in the vicinity of the crossing, and that it comes nearer to 
the surface in that vicinity than at some other places. 

South of the Croton there are several small cross faults run- 


ning nearly east and west. It is believed that similar movements 
have affected the rock in the Croton valley itself, modifying its con- 
dition so much as to control the course of the stream. The only 
immediate bearing upon the problem of the Croton crossing is the 
question that it raises about the quality of rock and the necessity 
that is introduced of trying to determine whether or not there is 
shattering enough to be very objectionable. 

Explorations and data 

Six drill holes have been made on this proposed Croton lake 
crossing — one on either side just at the margin and four others 
within the intermediate space of 1400 feet. These inner four have 
been made from rafts floated on the lake and have penetrated water, 
drift cover, and rock [see accompanying profile and cross section, 
pi. 27]. 

Rock floor. The depth of the preglacial Croton valley is 
pretty accurately determined at o feet or sea level. There is no 
reason to expect a gorge or inner channel of any consequence. 

The drills have penetrated only one formation, i. e. Manhattan 
schist. These test holes are believed to be near enough together to 
eliminate the possibility of any other formation appearing at tunnel 

Rock condition. The two varieties of schist (1) the coarse 
garnetiferous quartz-mica rock, which is a metamorphosed former 
sediment, and (2) the darker, close grained hornblendic rock that 
is believed to represent an igneous intrusion, both occur in the cores 
brought up by the drill. Either under normal conditions is a 
good rock. But there are considerable differences in the physical 
condition of the rock. Holes no. 3 and no. 4 at the two extremes, 
on the lake borders, show sound rock that comes up in large cores 
with very high percentage recovery. This is confidently believed to 
represent the average condition of the rock in this vicinity at the 
sides of the valley. 

The central holes, however, nos. 1, 2, 5 and 15, all show more 
broken ground. Of these holes no. 2 is much the most broken, the 
core recovery being only 14.8^. The pieces are small and many 
are smoothed (slickensided) by movement. The hole penetrates a 
typical crush zone resulting from slight faulting movements, and 
the low saving is due to the fact that the incipient fractures are not 
well bound together (rehealed) by later mineral change. They are 
probably connected with the latest movements of this kind. 


The commonest secondary mineral now filling these crevices is 
chlorite, and, although it may completely fill the crevices it has little 
binding strength. Any new disturbance or strain readily causes 
separation along the same original lines. But in spite of the fact 
that the core is broken into small pieces and shows so low percent- 
age of recovery it is quite certain that the rock itself is not badly 
decayed. An examination of one of the most doubtful looking 
cores from the lower part of hole no. x showed under the micro- 
scope little evidence of serious decay. This is believed to mean 
that underground water circulation is not as abundant as the 
fractured condition of the rock would lead one to expect. Further- 
more, an examination of the cores in greater detail shows beyond 
question that much of the fracturing is entirely fresh and must 
have been done by the drill itself. It is certain that the low per- 
centage of recovery is in part due to this cause. The small diam- 
eter of the intermediate holes is contributory to the same results. 
Some allowance must also be made for the difficulty of working 
a machine from a raft on the lake. 

Comparison of the cores shows a decidedly higher percentage 
of core recovery, and presumably therefore of rock solidity in all 
of the other three holes — no. i, no. 5 and no. 15. 

Hole no. 2 — core recovered 14.8^ 
" no. 1 — " 34-6$ 

" no. 15— 36.3^ 
" no. 5— " 38.9^ 

It therefore appears that the last three penetrate rock that is 
more than twice as good in its capacity to stand drilling disturbance. 

A comparison of quality at different depths is believed to be still 
more encouraging. The upper portions of all holes have poor 
recovery and comparatively poor looking rock. But in depth there 
is a marked improvement. 

In view of the fact that the tunnel will undoubtedly be located 
somewhere below the -75-foot level, it is really only this lower sec- 
tion that is of vital importance to the project. A tabulation and 
comparison of core recovery from these lower portions is given 

1 From total depth of hole 2 From depth -75' to bottom 
Hole no. 2 — = 14.8$ core recovery 25$ core recovery 
" no. 1— =34.6^ " 450 

" no. 15— =36.3-1! " 66> 

" no. 5— =38.9^ " 42^ 


Under the conditions of work, this is a fair saving and indicates 
much more substantial rock below the -75' level. There are many 
pieces 10-12 inches in length and for a 1 inch core this may be 
considered very good. 

It is clear, however, from a detailed inspection of the cores, that 
there is considerable variation somewhat independent of depth. 
There are occasional stretches of poorer ground in the midst of 
comparatively sound rock. This is believed to indicate that the 
crushed condition is confined chiefly to certain zones, and that these 
zones dip across the formation and across the holes at an angle. 
They are probably distributed promiscuously throughout the central 
portion of the valley, but are certainly more abundant and more 
strongly marked in the vicinity of hole no. 2 than at any other point 
tested. The rock profile shows that hole no. 2 has also the lowest 
bed rock. This is a further support to the general explanation of 
the valley as given above. 

The chief elements of uncertainty remaining after the borings 
have been completed are : 

1 The exact extent or widths of the chief crush zones 

2 Their dip and strike 

3 The possibility of others not yet touched 

4 The permeability of the rock for underground water 

5 The supporting strength of such rock in a tunnel of large 

In spite of the uncertainties enumerated, the conditions are 
entirely understandable. There is little probability of finding a 
worse condition than that shown in hole. no. 2. The permeability or 
porosity of these zones is of course unknown. The chief reason for 
believing that underground circulation is not abnormally heavy is 
the observation that the joints are well filled with chlorite and that 
other decay is not at all prominent at the lower levels. Further- 
more, the rock is a crystalline type of rather successful resistance 
to ordinary solution agencies and therefore may be depended upon 
to hold its own in its present condition indefinitely. But because 
of the poor binding effect of the chlorite it is to be expected that 
blocks will fall from the roof of any tunnel where it passes through 
a crush zone. Timbering will be required for protection in places, 
but the ground will not cave or run. These zones may be expected 
throughout a total distance of about 700 feet — i. e. the space 
between no. 1 and no. 15. The chief belt of such ground probably 
lies between holes no. 2 and no. 5. 



The lowest bed rock is about sea level. 

This pressure tunnel will cut only Manhattan schist. 

All rock is good ground for such work, except in certain narrow 
zones where it is crushed. 

The extent of such broken ground is not closely delimited, but 
occurs at intervals for a distance of 700 feet. 

The amount of underground circulation is judged to be moderate 
at -100 feet. 

The tunnel should be located deep enough to take advantage of 
the improved rock conditions shown at about -100 feet. There 
seems to be no marked improvement below -100 feet as deep as 
the drills have gone. 


Kensico reservoir at Valhalla, 2 miles north of White Plains, is 
one of the links in the Bronx river aqueduct. It is to be greatly 
enlarged and made a very important storage reservoir for the new 
Catskill system. In line with this plan a new dam is to be built 
near the old site that will rise 100 feet higher than the present 

Extensive investigations 1 have been made to determine the charac- 
ter of rock floor for this massive dam. Sites both above and below 
the present one have been studied with the question of safety and 
efficiency and permanence as well as that of economy of construc- 
tion in view. Involved with this is also the source of suitable stone 
for its construction. 

Geological surroundings 

Glacial drift covers the rock floor of this and neighboring valleys 
to a depth of 10 to 20 feet. No rock is exposed in the valley bottom 
at the Kensico site, but at the extremities of the proposed dam the 
rock floor comes to the surface in small outcrops. The material 
constituting the drift cover is essentially a loose, somewhat porous 
till passing into modified types, especially gravels and sands imme- 
diately south of the ground tested. 

The character of bed rock at the two extremities and beyond the 
limits of the dam is easily seen from the outcrops to be Fordham 
gneiss on the east and Manhattan schist on the west. Between, 
although nothing can be seen, Inwood limestone is found by the 
borings as was to be expected. No other formations occur, although 
the Yonkers gneiss, an intrusive in the Fordham at a little greater 
distance figures prominently in studies of material. 

The formations are in normal order and are of the usual petro- 
graphic character. All dip westward at angles that vary from 45 
to 65 degrees and have a general strike a little east of north. It is 
evident that the whole series represents an eroded limb of a simple 

1 These explorations have been in direct charge of Mr Wilson Fitch 
Smith, division engineer, whose headquarters for the Kensico division is at 
Valhalla, N. Y- Preparations for construction have already been begun. 




The Inwood limestone occupies about 800 feet of the bottom and 
eastern margin of the valley, lapping well up on the Fordham 
gneiss. The drill cores from this formation are unusually sound. 

The Manhattan schist shows much broken material. There are 
many crush zones. This condition increases still farther west along 
the railway near Valhalla station. 

The Fordham gneiss appears to be sound where it can be seen 
at the surface. 

Results of exploration. .Many borings have been made. They 
prove the general structure and succession of formations, making 
the boundaries definite. They increase the evidences of a rather 
wide prevalence of weak zones — some of them in the gneisses. 
And they also indicate a more extensive surface decay than was 
formerly believed to prevail. 

The chief problems from the geologic standpoint are connected 
with the following features : 

1 Extent of surface disintegration 

2 Extent and distribution of weak zones 

3 Depth of decay and perviousness of rock 

Surface disintegration. Several borings on ground underlain 
by Fordham gneiss penetrated material beneath the drift and above 
bed rock that was interpreted as residuary matter from rock decay. 
All of this material is of local origin. Later exploration in the 
form of a deep trench to bed rock has proven that there is an 
extensive residuary mantle of this sort at the eastern side of the 
valley below the present dam. In places as much as 30 feet exists. 
Undoubtedly this material is a remnant of preglacial soil mantle 
that was in some way protected from removal by the ice. Few 
places are to be seen in all southeastern Xew York where there is 
so much left in place. In most of it the gneissic structure is still 
preserved, but the decay is so complete that it can be cut and 
handled like an impure clay. 

Weak zones. It has been proven that there are weak zones 
in the gneisses as well as in the other rock formations. In some 
places the rock is so poor that no core is recovered for distances 
of 5 to to feet, and in one hole a seam of this kind 20 feet wide 
appears. In every case, however, the drill passes through the rot- 
ten material into the opposite wall — indicating a zone of consider- 
able dip instead of vertical position. This favors the theory that 
the weaknesses follow the bedding largely and are perhaps due to 



difference in the mineral make-up of 
the beds fully as much as to dynamic 
disturbances. The walls are generally 
good. The fragments of core are not 
much slickensided. In the schist this 
is probably not as generally true. 
There are much plainer evidences of 
crushing movements in the schist. It 
is a locality where (me of the folds, 
one well developed farther south, is 
pinched out and there is rather gen- 
eral crushing of the weaker strata. 

Depth of decay and perviousness. 
As deep as borings have gone there 
is occasional decay and broken ma- 
terial and streaks that are pervious. 

Final location. 1 "he condition of 
bed rock, together with other consid- 
erations led finally to the selection of 
a site above the present dam. In 
general the same features character- 
ize this site. But the rock condition 
is somewhat improved. On the whole 
the new situation is a safer one. 

3 i l O 




The following quarries in the immediate vicinity of Kensico res- 
ervoir have been studied in the field : 

(i) " Smith quarry," which is less than a thousand feet east of 
the southern end of the present reservoir; (2) " City quarry," which 
is on the immediate eastern margin of the reservoir on the east side ; 
(3) " Garden quarry," which is a new location about 500 feet from 
the eastern margin about midway; (4) " Outlet quarry," 1500 feet 
east of the northern extremity of the present reservoir; (5) " Ferris 
quarries " 1000 feet and (6) " Dinnan quarry " 3000 feet farther 

In addition to the field observations a detailed microscopic study 
was made on specimens of the rock taken from the Garden, Ferris 
and Dinnan quarries. 

The question at issue is the choice of a rock for the facing and 
finish of the new Kensico dam. In view of the use to be made of 
the rock, extreme strength is of only secondary importance. But the 
questions of abundance, distribution, durability, purity, agreeable 
appearance and working quality are vital. 

Types of rocks 

All of the quarries occur in the broad belt of Precambric gneisses 
that forms the eastern margin of the reservoir extending northward 
and southward for many miles. The formation as a whole is very 
complex. But the basis of it is a black and white banded rock 
chiefly a metamorphosed sediment, known as the Fordham gneiss 
in southeastern New York. In it are intrusions of igneous rocks 
of many varieties and most complicated structure — dykes, bosses, 
veinlets, stringers etc., sometimes in such abundance as to wholly 
obscure the original type. The most abundant of these are, (a) a 
rather light colored quite acid rock that is essentially a granite in 
composition, but has a sufficiently foliate structure to be classed as 
a gneiss and is the same as the " Yonkers gneiss " occurring farther 
south, and (b) a dark rock containing much hornblende and biotite 
which is in some cases essentially a diorite in composition, but has a 
marked tendency to schistose structure. The former (a) may be 
called a granite gneiss and the more massive representatives of the 
latter (b) may be classed as a dioritic gneiss. In both cases at 

7 I9h 



times the blending with the original metamorphosed Fordham gneiss 
is so intimate that absolutely sharp limits can not be drawn. And 
this last condition may well be designated as a third case (c). 

The quarries visited represent all three of these cases. Dinnan, 
Ferris and Outlet quarries represent essentially the " Yonkers 
gneiss" type (a) of granite gneiss. Garden quarry represents 
chiefly (b) the dioritic type of gneiss. Gty and Smith quarries 
represent the last case (c), or the mixed and variable type. 

Field character 

City quarry. In accord with the above differences in type it 
is found that large quantities of uniform material for such purpose 
as is proposed can not be obtained from City quarry. The rock 
there is badly jointed and is variable to a marked degree. It was 
not thought promising enough to test in detail. 

Smith quarry. The conditions of Smith quarry are better but 
there are similar objections. The amount of uniform material is 
greater. It would no doubt furnish an abundance of material suit- 
able for use in the construction of the dam interior, but is not at 
this point as. good a source of facing stone as some of the others 
to be considered. 

Outlet quarry. Although this rock is characteristic Yonkers 
gneiss, it has at this place suffered by weathering a peculiar dis- 
coloration to such extent as to make it objectionable, both from the 
standpoint of appearance and perhaps of durability. 

Garden quarry. There is an abundance of stone at the Garden 
quarry. It is fairly uniform. It is no doubt good enough from 
every standpoint of durability. It is well located. It can be quar- 
ried readily. But it has a very dark color and is undoubtedly less 
attractive than a light stone for this purpose. There are no objec- 
tionable structures, except where the strong schistose character is 
developed, and these could be avoided so that with a little selection 
a fairly uniform stone could be secured. 

Dinnan quarry. This rock is typical " Yonkers gneiss." 
There is sufficiently large quantity. It is of good quality. It is 
situated a little over 2 miles from the proposed dam, but is of easy 
access. The jointing and other structures do not seem to be objec- 
tionable. It will work somewhat more easily than a true granite 
because of the gneissic structure and it has a good medium light 
color. The discolorations do not seem to penetrate deep and the 
rock shows only slight decay. 

Plate 28 

Photomicrograph of Yonkers gneiss from " Outlet quarry " taken in 
plain light to show prominence of sutures between the grains indicating 
the beginning stage of disintegration. Magnified about 30 diameters 


Ferris quarries. The "Old Ferris quarry — is " Yonkers 
gneiss " considerably more weathered than the Dinnan. It is con- 
sidered less promising than the " New Ferries " quarry which has 
been explored by the engineers of the Kensico division. The rock 
of this quarry site is not all of one quality. There are essentially 
three varietal facies of the Yonkers gneiss type and relationship. 
One (a) is essentially a granite. It has a coarse grain and shows 
almost no foliate structure. It has a decidedly massive appearance ; 
but it is not of very great extent. This rock is evidently very 
closely related to the true Yonkers gneiss into which it passes on 
all sides through an intermediate variety. 

This intermediate variety (b) has medium size of grain, is only 
slightly foliated and passes without sharp limitations on the one 
side into the granite facies and on the other to true normal Yonkers 
gneiss. It is not so strikingly massive as the granite, but is more 
so than the gneiss proper. This rock may be called a gneissoid 
granite to distinguish it from the other. 

The true Yonkers (c) gneiss surrounds these two special varie- 
ties. It is of finer grain than either of the others and is more 
strongly foliate and is strictly a granite gneiss. Varieties (a) and 
(b) occur as sort of a lens within the Yonkers gneiss. 

The extent of the granite as now uncovered at the site is be- 
lieved to represent its 'limits. The prospect of enlarging the area 
will not meet with much success. It is essentially a local develop- 
ment connected with the differentiation of the parent magma from 
which all three varieties were derived. It seems to have been the 
last of the three to solidify, and it has some of the characteristics 
of certain pegmatite lenses. 

Although this is certainly an attractive rock and one against 
which there is little ground for objection, it is reasonably certain 
that a sufficient quantity of this variety can not be obtained here 
for the whole proposed use. And the prospects are not good for 
locating another quarry of the same quality. 

The gneissoid granite (b) is of greater extent, in fact it will be 
found to encroach on the present area of the granite. It is as good 
rock and almost as attractive as the granite. 

The regular type of Yonkers gneiss such as that represented in 
the Dinnan quarry can be obtained in almost unlimited quantity, 
and, with the splendid showing that it makes in further examina- 
tion, it has come to be considered the best suited to the purposes 
of dam construction at Kensico. 


Petrographic character of the rocks 

This line of investigation is confined to four sets of samples. 
No. i The granite of the New Ferris quarry 

2 The gneissoid granite of the same quarry 

3 The Yonkers gneiss of Dinnan quarry 

4 The dioritic gneiss of Garden quarry 

1 Granite. The rock is coarse grained and well interlocked. 
The chief constituents are orthoclase, quartz and microcline. 
There are but small amounts of dark minerals, and there is not 

much decay. 

Both surface material and the drill core were examined. The 
deeper material shows a little calcite, that may be original, occur- 
ring in irregular grains. They do not seem to indicate decay. 
There is a little kaolin alteration of the feldspars, but not to a 
serious degree. There are no injurious impurities in the rock such 
as might cause rapid disintegration or discoloration. 

The rock is undoubtedly of good grade as to strength, composi- 
tion and durability. 

2 Gneissoid granite (Ferris quarry). The rock is of medium 
grain, containing quartz, the feldspars and a little mica. 

There is very little alteration, and no serious decay or injurious 
constituents. A small amount of seriate and calcite present are 
not considered of consequence, and as in the case of the granite, 
the calcite is believed to be original. 

The grains are intimately interlocked and the rock is certainly 
of good quality and very similar to the granite proper. 

3 Yonkers gneiss (Dinnan quarry). This rock is fine grained, 
and is composed of quartz, mica and the feldspars among which 
microcline is very abundant. 

The condition is good, — very little alteration, close structure, but 
with a little more granular appearance than any of the other types. 

It is a good rock and gives good durability tests. 

On badly weathered surfaces the Yonkers gneiss breaks up into 
a granular product like sand long before it decays to earthy matter. 
This seems to be caused by expansion and contraction of the dif- 
ferent constituents under changing weather conditions inducing a 
weakening of the sutures. Sometimes there is very little decay 
even along these sutures, but as they open slightly they become the 
channels for moisture and staining solutions. This makes the 
boundaries of the grains very well marked in weathered specimens. 

Plate 29 

Photomicrograph of diorite gneiss from " Garden quarry." Magnifica- 
tion 30 diameters. The constituents are hornblende, biotite, feldspars 
and quartz. 



Such incipient disintegration is common in the more even grained or 
granular varieties. ' 

The accompanying photomicrograph [pi. 28] is taken in plain 
light and shows the outlines of the grains due to this cause. 

4 Dioritic gneiss (Garden quarry). Rock is of medium grain 
and with a strong tendency to schistose or foliate structure. The 
dark grains are hornblende and biotite, the light grains are feldspars 
and quartz. 

The rock is fresh, durable and has no injurious constituents. It 
is good enough for the use in all respects, but has a dark color and 
is more strongly foliated than any of the others considered. 

It is evident from these observations that the rocks considered 
are all of suitable mineralogic character for the purposes of large 
dam construction. For very large quantities of material, however, 
it is probable that neither the coarse granite nor the gneissoid 
granite could be depended upon for uniform supply. The true 
regular Yonkers gneiss, however, is very abundant, and can be relied 
upon for indefinite amounts. The dioritic gneiss is also abundant. 
The immediate region is not capable of furnishing any better rock 
than those described above. 

Additional tests 

Some instructive tests were made by the Board of Water Supply 
under the direction of Mr J. L. Davis who has charge of the 
testing laboratories. A few of these applying to the rocks at 
Kensico are tabulated below. 

The tests cover : specific gravity, weight per cubic foot, porosity 
in per cent, ratio of absorption, per' cent water absorbed, ratio of 
drying 24 and 48 hours, retained water pounds per cubic foot 24 
and 48 hours. 

In the accompanying tabulation the terms used are subject to the 
following limitations as to definition: 

1 Ratio of absorption, sometimes called porosity, " is the ratio of 
the weight of water absorbed to the dry weight of the stone." 

2 Porosity gives " the actual percentage of the stone which is 
pore space." " The difference between the dry and saturated 
weights of the sample is multiplied by the specific gravity of the 
rock and the product added to the dry weight. This gives the 
weight the specimen would have provided it contained no pore 
spaces. The difference between the dry and saturated weights 



multiplied by the specific gravity of the rock is then divided by the 
above computed weight of the poreless specimen. This ratio ex- 
pressed as a percentage is the actual porosity. Expressed as a 
formula, the computation is as follows: 

(Saturated wt. — Dry wt.) S. G. 

— ■ — — = Porosity." 

(Saturated wt. — Dry wt.) S. G. -j- Dry weight 

3 Ratio of drying. An attempt has been made to determine the 
comparative and actual rates at which the saturated rocks give up 
the absorbed water under ordinary atmospheric conditions. " The 
ratio of drying was computed by dividing the weight of water 
lost during exposure by total weight absorbed. The weight of re- 
tained water was computed." The comparison is most useful in 
rocks of like petrographic general character. 

The other terms need no explanation. 






: absorp- 
er cent 



per cubic 



Ratio of 

Retained water 
pounds per 
cubic feet 

No. of s 

Ratio 01 
tion p 





Per cent 






Granite, Ferris 
quarry, core No. 461 




2 .66 
2 .65 

164.7 \ 
164.0 s 

. 26 



. 224 

. 210 

Gneissoid granite, 
Ferris quarry, 
core No. 468 





2 . 63 
2 .65 

161 .0 ] 

162.8 J 




. 146 

• 145 

Yonkers gneiss, 
Dinnan quarry 




1 .01 

2 . 64 
2 . 64 

163 .3 1 
161 .0 J 




• 057 

• 057 

Dioritic gneiss, 
Garden quarry, 
core No. 459 



. 42 


2 .86 

175-4 ! 
174.8 1 


62 .5 


• 137 

• 137 

Gneissoid granite, 
Ferris quarry, 




. 96 
2 . 5° 

2 .63 

162.5 1 
159-4 1 

1 .08 





Granite, Ferris 
quarry, surface 



1 . 1 2 

2 .63 

162.3 1 
167-3 1 

. 40 

70 . 

74 -o 

. 207 

. 180 

Mr Davis concludes from a careful analysis and interpretation of 
these tests that the Yonkers gneiss is of superior durability. 



Geologic conditions as shown by exploration for a proposed pres- 
sure tunnel 

Bryn Mawr is a railway station 2 miles northeast of Yonkers. 
The general features of the vicinity, its topography, succession ot 
formations and the boundaries are shown on the accompanying 
sketch map which is largely copied from United States Geological 
Survey Folio No. 83. The Southern aqueduct follows southward 
along a Manhattan schist ridge until, at a point a*bout a mile northeast 
of Bryn Mawr, a cross depression of so great width and depth is 
reached that some special means of crossing has to be devised. 
Near Bryn Mawr station a gneiss ridge rises and continues south- 
ward. The proposed line follows this ridge. 

Explorations have been made as usual by drilling to determine 
if possible whether or not a bed rock pressure tunnel is practicable. 

The following questions may be made to cover most of the 
practical issues of the study : 

1 What formations would the tunnel cut? 

2 Which of these would show most questionable ground? 

3 What portion of the line is regarded as most critical — whose 
development would show whether or not a tunnel is practicable ? 

4 What special conditions are shown by drill borings? 

5 What interpretation is to be placed on the peculiar results from 
hole no. 4 where there has been unusually great difficulty in drilling? 

6 What experiences in similar ground have a direct bearing on " 
this case? 


The formations that would be encountered in the Bryn Mawr 
siphon are : 

1 Manhattan schist (top), the usual micaceous type, also called 
1 Unison schist in United States Geological Survey Folio 83. 

2 Inwood limestone (middle), the usual coarsely crystalline dolo- 
mitic and micaceous type, also called " Stockbridge dolomite " in 
the Folio, same as " Tuckahoc marble," same as " Sing Sing 
marble," same as limestone at Kensico dam and also at Croton dam. 




3 Fordham gneiss (bottom), the usual black and white thinly 
banded type, a much folded and strongly metamorphosed rock, the 
oldest of all. 

4 Yonkers gneiss, the usual type, gneissoid biotite granite very 
uniform and granular. This formation is an igneous intrusive that 
cuts up through the Fordham gneiss and is therefore younger. 
Whether it is also younger than the limestone and schist is not 

5 Quartz veins and lenslike segregations of quartz, also pegma- 
titic streaks, are occasional occurrences in all of the formations. 
They are most abundant in the schist, but are seen also in the Ford- 
ham gneiss. A similar development was encountered in the lime- 
stone in hole no. 40. 

6 Glacial drift, chiefly modified drift, partially stratified sand 
and gravel, reaching more than 125 feet in depth, covers por- 
tions of all formations. 

This last formation (no. 6) is the only one that may be wholly 
avoided in the tunnel proper. The chief interest lies in its hindrance 
to exploration and its possible usefulness as a source of sand and 
gravel supply. 

Weakest formation. The Inwood limestone is the most ques- 
tionable ground. This is believed to be so chiefly because of the 
greater solubility of the rock, its granular and micaceous character, 
and the probability that a line of displacement accompanied by some 
fracturing crosses the siphon line in this formation. If a very 
excessive amount of shattering occurs in this zone it may have 
induced a condition of disintegration to such depth as to endanger 
the tunnel. 

There are no surface indications of a serious condition at depth 
for any of the other formations. 

Critical zone 

The critical zone is probably not far from the contact between 
gneiss and limestone. There are two reasons for this opinion. The 
first is related to the nature of the folding. The formations are 
squeezed into a close syncline pitching northward. In cross section 
the strata at any point around the head of this trough dip inward, 
and, because of the more resistant Fordham gneiss forming the floor 
of the trough, the drainage and seepage and consequent tendency to 
decay might be expected to follow along its upper contact. 

Plate 30 

n\:-0/t)frf. ftp**/? secTwd 

Location map showing by the dotted belts the distribution of Inwood lime- 
stone in the Hastings- Yonkers district and the position of the Bryn 
Mawr tunnel section as well as shaft 13 on the New Croton aqueduct 
with their relations to the limestone belts. Manhattan schist and Ford- 
nam gneiss occupy the rest of the area. 


The second reason is related to the probable later faulting move- 
ments. It is evident from the map [Folio 83] that the formations 
in the vicinity of Bryn Mawr are bulged up. One would expect 
the trough which contains the schist and limestone of Grassy Sprain 
valley to continue uninterruptedly southwestward and join with 
Tibbit brook valley. But a cross fold has bulged the formations 
up so much that for a distance of a mile erosion has removed all 
of the formations except the gneiss. Bryn Mawr station is about 
central on this bulge. Evidence of such a movement is readily 
seen on the gneiss along the northerly margin where it slopes down 
toward the limestone. The movement had developed a little shear- 
ing and has tilted the minor folds downward toward the north at 
angles varying from 30 to 8o° from the horizontal. This angle 
becomes somewhat more accentuated as the limestone is approached, 
and it is believed that it may pass a short distance into the limestone 
border. There is, however, no great amount of crushing evident in 
the gneiss and this may hold also in the limestone. 

The fact that Sprain brook crosses the formations along this 
northerly margin and flows for 2 miles in a southeasterly direction 
may indicate a still later movement, probably faulting. There is no 
surface evidence of it except the abnormal course of the creek. 
But, if there is such a fault, it also crosses the siphon line in the 
same zone, i. e. in the vicinity of the limestone-gneiss contact, not 
far from the location of the present course of the brook. 

Therefore it seems reasonable to conclude that the critical zone 
is near the contact, probably on the limestone side, and in the 
vicinity of the present course of Sprain brook. It is also probably 
cut deepest here by erosion. If this zone is in good enough condition 
to stand tunneling the rest of the line ought to be. 

Conditions indicated by borings 

All rock formations stand very steep. They vary from 8o° to 
90 . This means that very few beds can be explored by one hole, 
and that any weakness or crevice is likely to make a showing in 
excess of its true proportions. 

The cores show considerable crushing. Some of the fractures 
are not healed, although weathering from circulation is not present 
on all of them. The micaceous layers are most affected by circula- 
tion. Some beds of this variety are considerably weakened even at 
depths of over 200 feet. Occasional seams have been encountered 
that give no core at all for several (even 20 or 30) feet. But the 



greater proportion of the recovered pieces are comparatively solid 
even where the total percentage of saving is very low. It is evi- 
dent that some of the core, a considerable percentage, has been 
ground to pieces in the process of boring. This is especially notice- 
able at hole no. 40. 

Hole no. 40. Much trouble has been met in this hole. A 
careful analysis of the record and core and the behavior of the drill 
is interpreted as follows: 

1 Partially assorted drift, chiefly sand and gravel was penetrated 
for 125 feet. 

2 Limestone bed rock of fairly sound quality was struck at 
about 125 feet (about el. -40). 

3 The casing that was put down to shut out the sand failed to 
reach solid rock, and this permitted a continual supply of pebbles 
and sand to run into the hole and obstruct the work with each pull 
up. The presence of these pebbles was also instrumental in grinding 
the core to pieces, and this accounts chiefly for the low saving. 

4 After this opening was plugged up with cement, the drilling 
was continued successfully until a somewhat broken quartz vein 
was encountered and this has been followed for about 35 feet. Its 
broken condition afforded another opportunity for fragments to fall 
into the hole, and on top of the drill, bringing the work for a second 
time to a standstill. It is certain also that the drift pebbles still fall 
in. As the formation stands vertical here it is not surprising that 
any feature should show an apparent extent quite out of proportion 
to the real value. The quartz vein is probably of no great breadth. 
Small seams containing mud may also be followed 15 or 20 feet 
and still be of no great significance in the formation as a whole. 
The rock fragments (core) recovered in this hole are fairly sound. 

5 In spite of the many delays and difficulties of this hole, it is 
apparent that the general rock formation is not responsible for it 
all. The failure to reach solid rock contact with the casing has been 
the cause of part of it. Later the penetration of a rather rare 
quartz vein, a thing that would not often be found in the limestone, 
has added to the trouble. Both of these causes are so rare that 
they may almost be given the value of accidents. 

But the last 100 feet or more of the hole, from depth 225 feet to 
335 feet, shows an unusually questionable condition. Only a few 
rock fragments are saved and they include limestone and quartz 
vein matter. The rest is wholly disintegration sand of rather com- 
plex composition but carrying very much mica. This is all wash 




material except one sample, which is a " dry sample " and is still 
more strongly micaceous. 

Borings nos. 40, 45 and 46 are all within the zone that was con- 
sidered, from surface indications, to he likely to carry the deepest 
gorge and to show the weakest rock. Because of the heavy drift 
cover (more than a hundred feet) it is manifestly impossible to 
locate the weakest zone more closely or judge of its exact condi- 
tion except by borings. 

Hole no. 42 at station 634 + 28, penetrates 82.4 feet of drift and 
reaches bed rock at about elevation 21 feet A. T. The rock is good, 
substantial, coarsely crystalline limestone. It shows as sound con- 
dition as can be expected in this formation even under the most 
favorable situations. 

Hole no. 46 at station 644 + 77.4 is just south of the brook. It 
penetrates 72 feet of drift and reaches bed rock at elevation 14 
feet A.T. The rock is Fordham gneiss of typical sort and in per- 
fectly good condition. There is no question about the soundness of 
the rock from this point southward. 

Hole no. 45 at station 643 + 52.5, 125 feet north of hole no. 46 
penetrates drift for about 150 feet (possibly a few feet less, 145 
feet). This drift cover is interpreted as mostly sand (modified 
drift) to 115 feet and a boulder bed from 115 to 143 feet. After 
the true ledge is reached it is sound and shows no unusual or ques- 
tionable conditions. It is Fordham gneiss. 


1 Weak zone. There is little doubt that this last 100 feet of 
hole no. 40 is in the decayed weak zone that was expected to de- 
velop in the vicinity of the contact between the gneiss and the lime- 
stone. It would be expected to pitch northward along the floor of 
gneiss and extend beneath the southerly extremity of limestone at 
this point [see fig. 36]. 

2 Contact. Hole no. 40 cuts limestone, hole no. 45 cuts only 
gneiss, therefore the formational contact lies somewhere in this 
177-foot space. 

3 Position of old channel. Bed rock surface is lowest at hole 
no. 45. But since the rock itself is sound gneiss, it is not believed 
to represent the lowest possible point. This is still more certain 
because of the fact that the pitch is northward so that this becomes 
a dip slope on which the prcglacial stream could glide against the 
edges of the limestone beds [see diagram], and because the condi- 



tion of the rock a little farther north (at hole no. 40) shows that 
these limestone beds are actually much weaker than the gneiss. 
Therefore the deepest portion of the buried channel is to be expected 
between holes no. 40 and no. 45, and probably nearest to hole no. 40. 

4 Depth of old channel. How deep the buried channel may 
be can not be accurately estimated. But if the same dip slope as 
is shown by the rock surface from hole no. 46 to no. 45 prevails 
northward toward hole no. 40, a depth somewhat below -100 feet 
may reasonably be expected. In the absence of data bearing upon 
the depth of other portions of this ancient channel or of the lower 
Bronx river with which it must have been connected, it is impossible 
to estimate more closely. 

5 Interpretation of hole no. 40. There is so little rock actually 
saved from the more than 200 feet of possible core on this hole 
that its real character is very obscure. 

There are three possible explanations for the condition found in 
the last 100 feet. 

a The drill may have followed a large mud seam. 

b The material may be only residuary rotten limestone still wholly 
above the gneiss. 

c The actual contact may have been penetrated and a part of 
this rotten material may be decayed gneiss within a crush zone. 

The difficulty in drawing absolute conclusions is increased by the 
fact that matter falling in from above has been a continued source 
of trouble and is more or less mixed with the rock material of 
lower points. Therefore, the fact that the sand taken from the 
lowest points, 335 feet, is silicious instead of calcareous, may not 
prove satisfactorily that the rock at that point is wholly silicious. 

It is worth noting, however, that the harder rock in the upper 
portion of the hole was in places much crushed and that mud seams 
were encountered before reaching this last 100 feet. 

It is also worth noting that the same dip slope of rock surface 
as prevails between holes no. 46 and no. 45 if continued northward 
to hole no. 40, would cut that hole a considerable distance (75 feet) 
above its bottom. 

In view of all the conditions, therefore, it is judged that there is 
a crush zone here, that hole no. 40 penetrates it, that it is badly de- 
cayed, that the plane of the crush zone dips steeply northward and 
cuts both limestone and gneiss, that a tunnel at about -300 feet 
would cut this zone south of station 640 and north of station 642, 
and that all other portions of the line are in comparatively satisfac- 


tory condition. This zone for a hundred feet is likely to be wet, 
weak, and would require extra precautions and additional expense 
in construction. 

6 Evidence of faulting. Whichever interpretation of hole no. 
40 is taken is in support of some displacement in the nature of 
faulting between holes no. 40 and no. 45. If the gneiss rock floor 
is not reached in hole no. 40, then the greater northward slope of 
it from hole no. 45 to no. 40 than is shown from no. 46 to no. 45 
indicates a downward movement. If on the other hand, the iden- 
tity of the formation in the lower part of hole no. 40 be considered 
undetermined, and its condition attributed to decay in a crush zone, 
the presence of the crush zone itself indicates movement of a fault 

Conclusions as to character of the crossing 

In considering the geological conditions as a factor in the prob- 
lem of practicability of a tunnel, it is necessary to note the follow- 
ing points : 

1 In view of the fact that the deepest point in the ancient chan- 
nel is not yet found, and that it will probably go below -100 feet, 
it would be necessary to figure on a tunnel grade down well toward 
-300 feet. 

2 It would be necessary to figure on a wet and weak zone of at 
least 100 feet along the tunnel and a more expensive construction 
at that point. 

3 The ground at such depth south of station 642 is unusually 
sound. The ground north of station 636 may be counted good. 
The ground between 636 and 640 may be considered fair, and the 
ground from 640 to 642 +, troublesome, containing the chief ele- 
ments of uncertainty. 

Fig. 36, which is a geologic section along the line at this point, 
shows the distribution of these features drawn to scale. 




[Sec outline location map, pi. 30] 

There has been reference made occasionally in connection with 
the Bryn Mawr explorations, as well as others, to the remarkable 
piece of bad ground encountered in 1885 on tne New Croton aque- 
duct near Woodlawn in the Saw Mill valley. This experience has 
been the source of much misgiving. Because of its evident im- 
portance and close relationship to conditions that may exist in the 
same formation at points on the Catskill line, an examination of 
this ground was made for the purpose of comparison. The mean- 
ing of that case and its bearing on the Bryn Mawr questions a~e 
given below : 

Engineer's records 

This ground and its remarkable behavior is described by Mr J. P. 
Carson in the Transactions of the American Institute of Mining 
Engineers, September 1890, pages 705-16 and 732-52. 

A description is also given in Wegman's Water Supply of the 
City of New York, 1658 to 1895, on page 152. 

From Mr Wegman's report is taken the following: 

The south heading was started from this shaft on June 1, 1885. 
It advanced at the rate of about 80 feet per month for 392 feet 
through good limestone rock (dolomite), which then became softer. 
On December 9, 1885, when the heading had reached a point 407 
feet from the shaft a fissure was encountered from which about 
100 cubic yards of decomposed limestone clay, sand and dirty water 
poured into the tunnel, partly filling it for a distance of 125 feet. 
After three days delay, when, only clear water was flowing into the 
tunnel, the fissure was plugged with straw. The heading was ad- 
vanced 20 feet further until on December 22, 1885, an outpour three 
times greater than the first occurred, covering everything in the 
heading out of sight * * * borings were made on the surface 
with a diamond drill to determine the extent of the soft ground in 
front of the tunnel. It was found to lie in a pocket in the rock, 




which had a length of no feet on the axis of the tunnel and ex- 
tended for a short distance below the invert of the conduit. The 
soft material, consisting of sand, gravel, clay and decomposed rock 
had a depth of about 160 feet from the surface to the top of the 
tunnel. It exerted such a pressure against the timber bulkhead that 
the 24-inch oak logs used as " rakers " (braces) became crushed in 
24 hours and had to be continually renewed. 

The chief points of present interest are that the tunnel, at a depth 
of about 160 feet from the surface, and after passing through sev- 
eral hundred feet (407 feet) of good dolomite, came into rotten rock 
and soft ground no feet across on the line. It was so soft that 
it ran into the tunnel in great quantities and exerted such pressure 
as to make progress in it a very troublesome and costly matter, 
taking " 60 weeks to advance the tunnel 85 feet " and costing " $539 
per foot." The material caved in so freely as to form a pit on the 

Statement of geologic conditions 

It is not possible to interpret the conditions at this locality as 
fully as one would wish because of the vagueness of some of the 
statements, but the following facts and explanation are essentially 
correct : 

1 The rock is the Inwood limestone, the same kind and same 
general conditions as all of the limestone belts that occur in the 
region of the Southern aqueduct. 

2 The soft ground penetrated at the point in question — 407 feet 
south of shaft 13 — called in the Carson report and others " a fis- 
sure " or " pocket," etc., is in reality a fault crush zone. The fault 
plane probably dips steeply southeast and strikes n. 50 e. cutting 
the tunnel line at an angle of something like 20°. 

3 The point is well up on the side of the valley more than a 
hundred feet above Saw Mill river, and the strike of the fault zone 
in its southwesterly extension cuts into the lower portion of the 
valley, so that underground circulation would be encouraged along 
the zone in this direction. 

4 The limestone outcrops very near by on the west side of the 
line and the Manhattan schist occurs near by on the east. The atti- 
tude of the beds is such as to indicate a fault of the thrust type, 
The accompanying figure illustrates this relationship in a cross sec- 
tion at right angles to the axis of the tunnel [see fig. 37] - 


Cvubh ~Z-o no. S 

Fig. 37 Sketch of the geologic structure at shaft 13 on the New Croton aqueduct. 
Interpreted from field observations 

5 It would appear probable that this zone was penetrated at the 
worst possible level, i. e. near enough to its Wholly decayed upper 
part to furnish no resistance at all to the overlying sand and gravel, 
and not deep enough to reach the more substantial (although prob- 
ably crushed) rock that may reasonably be expected to prevail at 
no very much greater depth. 

The chief point is that the weak spot has a reason and is not an 
accidental thing that might be expected just anywhere. But it must 
be admitted, in spite of this fact, that a casual examination of the 
locality would not make one suspicious of its existence, and it is 
surprising that the spot could have caused so much trouble. 

From the above it will be seen that in several respects the Bryn 
Mawr case is somewhat similar to this. They both indicate fault- 
ing ; they are in the same type of rock ; they both show or indicate 
caving tendencies. 

On the other hand, there are certain elements of difference some 
of which are capable of very materially modifying any conclusion 
that might be based upon the simple facts of likeness. For exam- 
ple — it should be expected (i) that the fault movement at shaft 13 
would be the greater because of lying in the more prominent lines 
of such displacement of the region, (2) being a thrust movement, 
the crush effect is probably more prominent at shaft 13 than at 
Bryn Mawr, (3) occurring at greater elevation above probable cir- 
culation outlet, the opportunity at shaft 13 for extensive and rather 
deep decay is the greater, (4) being cut so near the surface (160 
feet), its condition there is not necessarily a reliable guide to the 
seriousness of decay at a greater depth. 



Comparison of Bryn Mawr and shaft 13 

The following statements embody an opinion on the points raised 
or suggested in connection with a reference to the New Croton 
difficulties at shaft 13. The items are therefore treated by compari- 
son or contrast so far as possible : 

1 Type of rock. The rock explored at the Bryn Mawr siphon 
is the same formation as that in the Saw Mill valley cut by the 
New Croton aqueduct, i. e. the Inwood limestone — sometimes 
called " Stockbridge dolomite." It is the same also as the other 
large limestone belts in Westchester county. There are occasional 
small strips of limestone of another type, but its behavior could not 
be very different. 

2 Soft material. " Is any material of this sort " (like that in 
the New Croton tunnel near shaft 13) " likely to be encountered 
either in the crushed zone at boring 40 or elsewhere in the lime- 
stone belt? " 

It is sure to be encountered, especially near hole 40, if that zone 
is cut shallow. The behavior of the lower portion of this hole is 
very similar to the described case near shaft 13. The only prob- 
ability of avoiding it lies in placing the tunnel deep enough to cut 
more substantial rock. The single hole upon which all this argu- 
ment is based can scarcely be considered a thorough enough ex- 
ploration to build up a quantitative statement as to depth or width. 

There is no evidence, either on surface or in the exploration 
holes, of any other such zone on this line. 

3 Depth and extent. Under the circumstances, the increased 
depth makes it less probable that so much ground of like behavior 
would be found. Again, it is not likely that precisely the same 
conditions would so effectually halt operations or be considered so 
nearly insurmountable at this time. One of the many serious 
objections is that the tunnel would have little strength or resist- 
ance to a bursting pressure. It must be admitted that if caving 
ground were penetrated it would prove very difficult to handle with 
the gravel cover at the depths now considered, i. e. 300 feet or 
more below the surface. 

4 Water. " AVhat are the probabilities in regard to the quan- 
tity of water to be met in the crushed zone near boring 40? Can 
any limit be set which it would be extremely improbable that the 
inflow would exceed, on account of the topography of the country 
and the nature of the overlying materials?" 

There is likely to be much water. Nearly all of the overlying 



drift is sand and gravel that is probably saturated and in such con- 
dition as to permit easy flow to any lower outlet. It may readily 
carry 8-10 quarts of water to the cubic foot or about 2 gallons. 
The area covered by such deposits is about 2500 feet long on the 
southerly base along the creek and at this margin is approximately 
150 feet deep. The northerly margin is variable and reduces in 
places to o feet in thickness. It may, however, really represent 
500,000,000 cubic feet of this gravelly material holding 1,000,- 
000,000 gallons of water as a nearly permanent supply. 

This overlying material is necessarily a menace of no mean pro- 
portions. Every crevice or crush zone remaining unhealed will 
have water and plenty of it, the inflow being limited only by the 
size of the cracks and their abundance until the reservoir should 
be drained. There is no hardpan bottom to act as a dam. 

Outside additions to this permanent supply are confined to that 
received from rain and the stream. The rainfall on the area and 
immediately available as addition to the underground supply in the 
lower sands, together with the stream flow, which would probably 
sink into the sands, if an attempt to drain the underground supply 
were made, may be expected to furnish additional water at a pos- 
sible rate of 2500 gallons per minute. How much of all this is 
available at tunnel level depends wholly upon the openness of 
structure in the rock. There is nothing else to materially control 
the permanent and additional supply. 

There is evidence in hole 40 of considerable crushing. That 
means capacity for water circulation, but how much no one can 
tell. There is also much rotten rock in the same hole. This means 
that circulation has been easy and effective, but how much now no 
one can tell. The single hole (no. 40) in the absence of any other 
corroborative data is not sufficient to base more elaborate or precise 
quantitative estimates upon. 

5 Solubility. What is " the nature of the limestone with 
reference to its resistance to solution?" 

This limestone is, as all limestones are, more easily attacked by 
circulating water than most other rock types [see Rondout Valley]. 
The Inwood limestone such as occurs at Bryn Mawr is crystalline, 
often contains much mica and then is inclined to be foliated in 
structure, and it prevailingly stands steeply inclined. Because of 
these features in which it differs from the Rondout Valley lime- 
stones, it is likely to be more generally affected by decay along the 
zones permitting circulation than any of the Rondout Valley types. 



The Rondout Valley limestones are affected along joint planes, but 
the effect is almost wholly confined to a simple enlargement of these 
crevices. In the Inwood an additional effect is the weakening of 
the sutures or bond between the individual granules resulting in a 
tendency to weaken the whole mass as far as there is much pene- 
tration of seeping water. It would have less tendency to produce 
openings or caves, but greater tendency to produce a rock that 
would crumble in the hand or that would gradually assume the con- 
dition of a lime sand or a micaceous mud. 

As to the effect of water from the aqueduct on fresh portions of 
this rock, it is certain that the rock would be attacked wherever 
exposed to direct action. Its method of attack is by solution, and 
the rate of attack may safely be reckoned as not materially different 
from that assumed or being established by experiment and experi- 
ence on the Rondout V alley types. 

In the final consideration of the difficulties at Bryn Mawr the 
engineers have decided to abandon the tunnel plan. It is probable 
therefore that no additional explorations of direct bearing on the 
problems of this ground will be made. 



Hill View reservoir is the terminus of the Southern aqueduct. 
The Catskill water is to be delivered at this point, just north of the 
New York city line on the Yonkers side, at an elevation of 295 
feet. From this reservoir the water is to be distributed by an inde- 
pendent system of conduits to the principal centers of consumption 
in lower Manhattan and Brooklyn. 

It is believed that distribution can be most economically made 
and the system be most permanently established by constructing 
the main trunk distributaries as tunnels in solid bed rock at con- 
siderable depth below all surface disturbances. 

Preliminary investigations have been carried on by Headquarters 
department, Mr Alfred D. Flinn, department engineer, beginning 
in 1908. As the active work of exploration was entered upon Mr 
William W. Brush, department engineer, was assigned to this special 
division of the department's work and most of the preliminary ex- 
ploration borings were planned and finished under his immediate 
supervision. With the resignation of Mr Brush to take the post 
of deputy chief engineer in the Department of Water Supply, Gas 
and Electricity, Mr Walter E. Spear, department engineer, was 
secured to continue the difficult work of finishing explorations and 
preparing for construction. 

Studies of conditions affecting such a system and explorations 
designed to test the ground in line with these studies 1 have been 
made. The work thus far done in an exploratory way has been 
confined to one main distributary. 

Section A. Preliminary geological study 

As a preliminary step toward the systematic study of local con- 
ditions affecting possible conduits, trial lines were laid out on the 

x Few engineering enterprises, probably, have been planned with so care- 
ful regard for all known geologic conditions. The geologist and the en- 
gineer worked alternately on the same problems until, in the opinion of both, 
the best possible line was selected. It is the writer's belief that so sys- 
tematic a method has seldom if ever been carried out in engineering work 
of this kind. On this account, and in part to illustrate some of the pre- 
liminary stages in such work, many of the original facts and arguments 
and suggestions are given without change in the following discussion. 




city map from Hill View reservoir to Brooklyn by three different 
routes. So far as the topography and city development and other 
engineering considerations could be forseen either route could be 
u<ed. Studies of all kinds were expected to indicate which would be 
the most favorable and whether or not it might be advisable to shift 
even the best one to still more favorable ground. These are shown 
on the accompanying map which also covers the local geology of 
the immediate vicinity of the lines [see pi. 32]. 

General questions 

When the problem of the practicability of a rock tunnel for 
distribution conduits was first studied, several general questions 
were raised which indicate the lines of investigation followed. 

1 What is the character of the rock along the projected conduit 
lines shown at the depths required for such tunnels? 

2 Will the rock at moderate depths be such as to permit success- 
ful and economical construction of tunnels to be used under the 
hydraulic pressure due to Hill View reservoir? 

3 Does the character of rock in the vicinity of the lines vary 
sufficiently to materially affect the cost of a tunnel if the lines be 
shifted approximately 1000 feet either way from those shown on 
the original map as trial lines? 

4 Are the suggested locations of conduit lines adapted from a 
geological viewpoint to the construction of pressure tunnel con- 
duits, and, if not, what changes in these lines would be advisable ? 

5 Is the thickness of rock covering sufficient at all points to 
obviate trouble from open seams and disturbed surface rock? 

6 What borings and other field investigations should be under- 
taken to determine the practicability of construction of pressure 
tunnels along the lines suggested? 

In line with this series of questions a thorough geological investi- 
gation was begun, the chief conclusions of which are given below. 

Geological formations 

There are six local formations of sufficient permanence and in- 
dividuality of character and of sufficient areal importance to be 
treated as units in this study. These are described in some detail 
in part 1, but for convenience are briefly listed as follows: 

1 Glacial and postglacial deposits of boulders, clay and sand, with 
silt beneath the rivers. 

Plate 31 

A relief map of New York city and environs. Reproduced from a model 



2 Manhattan schist, the most abundant formation, chiefly mica 
schist with very subordinate hornblende schists, and usually with 
abundant pegmatite lenses and veins. 

3 The Inwood limestone, a white, dolomitic marble when fresh, 
which shades into impure, micaceous varieties. 

4 The Fordham gneiss, varying from a thinly schistose or 
quartzose rock to a strongly banded or a very massive and much 
contorted gneiss. The oldest formation of the district. 

5 The Yonkers gneiss, an original intrusive granite, now 
squeezed into a gneiss. Younger than the original Fordham. 

6 The Ravenswood grano-diorite or as it might be called in 
engineering practice, granite ; an original, intrusive rock now some- 
what gneissoid from pressure. Younger than the original Fordham. 

The Manhattan schist, the Inwood limestone and the Fordham 
gueiss are cut by veins or dikes of coarsely crystalline granite, 
technically called pegmatite. They are of irregular distribution and 
do not affect the tunneling operations one way or another. 

All the formations older than the glacial drift have been com- 
pressed into a series of northeast and southwest folds, and all have 
as a rule a steep or almost vertical dip. The axes of the folds are 
not horizontal, but usually pitch downward to the south at low 
angles. Erosion has developed a series of ridges trending north- 
east and southwest. The limestone being a softer and more easily 
eroded rock, almost always underlies the valleys or flats and the 
river channels. It is certain also that there is some faulting. 

Rock at depth 

The distribution of geological formations along the proposed 
lines has been shown on the accompanying map [pi. 32]. In gen- 
eral the kind of rock at tunnel depth will be the same as at the 
surface as indicated on the map for each point. Such error as there 
is, arises from two causes : (a) Uncertainty as to the exact location 
of some of the contact lines between two formations (usually due to 
drift cover), and (b) dip and pitch of the strata. 

In the first case (a) where the drift is particularly heavy, it is 
sometimes impossible to fix a contact line accurately from surface 
features alone. 

In the second case (b) it must be appreciated that nearly all of 
the formations dip eastward at a very steep angle, so that a form- 
ation would usually be found to extend a little further east at depth 
than at the surface. And also all formations pitch southward, so 



that they would be found to extend considerably farther south at 
depth than their surface outcrops. This angle of pitch is from 
io° to 30 . 

In nearly all these cases, however, the obscurity of the actual 
surface boundaries is as great a source of uncertainty as the effect 
of dip and pitch, so that the boundaries as mapped may be con- 
sidered sufficiently accurate for this comparative study of the lines. 

It is worth noting that the rock at the proposed depths of tunnels 
would be, as a rule, more substantial than at the surface. But there 
are several places on all of the lines where the exact condition is 
unknown at the surface as well as at depth. The chief points of 
this character will be noted in a later paragraph. 

Comparison of lines 1 

A comparison of the three lines submitted as the basis of ex- 
amination — (a) the westerly one, (b) the central one, (c) the 
easterly one [see accompanying map, pi. 32], as to rock formations 
likely to be cut by them, furnishes the following figures : 

Line A. Going southtvard from Hill View reservoir 


6 200 Yonkers gneiss — good rock 
1 400 Fordham gneiss 

1 400 Probably largely Inwood limestone with one weak zone 

(at Van Cortlandt lake) 
5 600 Fordham gneiss — good rock 

2 400 Near contact with limestone, probably in gneiss 
1 600 Crossing Harlem river — Inwood limestone 

4 000 Inwood limestone — probably fairly good rock 
800 Inwood limestone - — probably containing bad zone to 

16400 Manhattan schist (to 135th st.) 
2- 000 Along contact between schist and limestone 
4200 Inwood limestone with one weak zone (to s. end of 
Morningside Park) 

1 The statements of quality and extent of certain formations and zones 
are capable of some modification as exploratory work progresses. Some of 
these are noted in later sections of this report under special headings, such 
as The Lower East Side, and The East River-Brooklyn section. For the 
present purpose, as showing the development of the geologic basis of the 
project it seems preferable to leave the accompanying comparisons in their 
original form as presented to the board. 



12 800 Manhattan schist probably good quality (to s. end of Cen- 
tral Park) 

21000 From Central Park to East river — no outcrops — mostly 
Manhattan schists at tunnel depth. Condition largely 
conjectural 1 — probably mostly good rock with occasional 
weak zones 

6000 Manhattan island to City Hall, Brooklyn. Containing an 
unknown 1 zone in the East river and unknown quality 
of rock in Brooklyn. 

Summary of Line A 


6 200 Yonkers gneiss 

7000 Fordham gneiss 

2400 Contact (probably in gneiss) 
12000 In wood limestone 

2000 Contact ( probably in limestone) 
29200 Manhattan schist (good) 
21 000 Estimated Manhattan schist (fair) 

6000 Almost unknown 

85 800 total 

Line B. Going southward from Hill View reservoir 


8 000 Yonkers gneiss — good quality 
13000 Fordham gneiss — good quality 
6 800 Inwood limestone, probably mostly in fair condition, except 

at two points (to Cromwell av.) 
6600 Inwood limestone, unknown condition, but probably largely 
poor (to Harlem river) 
600 Inwood limestone — unknown condition (Harlem river) 
4600 Inwood limestone — unknown condition — probably fair 
(to Mt Morris Park) 
800 Manhattan schist, good 
800 Probably Manhattan schist — unknown 
2 800 Inwood limestone — unknown condition ■ — probably at least 

one bad zone (to 106th st.) 
12000 Manhattan schist along Central Park — good 

1 Explorations since conducted by the Board of Water Supply have proven 
the quality and character of the rock floor at these places. For the revised 
statement on these sections see the special discussions. 




8600 To Broadway — Manhattan schist (little known except 

from tunnels already made) 
14000 To East river, prohobly Manhattan schist (same as line A) 
6000 Manhattan island to City Hall, Brooklyn — uncertain con- 
dition (same as on line A) 



8000 Yonkers gneiss — good quality 

13000 Fordham gneiss — good quality 

21400 Inwood limestone — variable quality 

12800 Manhattan schist — good quality 

23 400 Estimated Manhattan schist — fair 

6 000 Almost unknown 

84 600 total 

Line C. Going south from Hill View reservoir 


6 000 Yonkers gneiss — good rock 

17400 To Webster av. — Fordham gneiss — good rock 

5 000 Along contact between limestone and gneiss 

9800 To 138th st. — Inwood limestone with probably two bad 

1 800 To Bronx kills — along contact between limestone and 
gneiss — uncertain quality 
600 Across Bronx kills — mostly in limestone containing a 
fault zone — probably bad ground 

6 400 Crossing Randall's and Ward's islands and Little Hell Gate. 

Nearly all is Manhattan schist of good quality 
1 000 Crossing Hell Gate — Inwood limestone 
1 200 Crossing Hell Gate — Fordham gneiss of good quality 
1 800 Astoria point — probably Fordham gneiss of good quality 
1 000 Crossing another limestone belt 

1 000 To Vernon av. — Fordham gneiss of unknown quality con- 

taining one fault zone 

7 000 To Nott av. — Ravenswood grano-diorite — good rock 

2 800 To Borden av. — Probably Ravenswood grano-diorite. 
18400 To Fort Greene Park Brooklyn — almost wholly unknown 

but contains probably 5000 or 6000 feet of poor ground 





6 000 Yonkers gneiss — good quality 
17400 Fordham gneiss — good quality 

6800 Along contact between limestone and gneiss (questionable) 
12400 Inwood limestone — with several bad zones 

6 400 Manhattan schist — probably good quality 

3 000 Fordham gneiss — probably good quality 

1 000 Fordham gneiss — unknown quality 

9 800 Ravenswood grano-diorite — mostly very good rock 
18 400 Almost wholly unknown 

81 200 total 

Tabulated summary — Types of rock formations 



line b (central) 














6 000 



13 OOO 

(IS- 3) 

21 400 


. , 4 4OO 



6 800 


12 000 


21 400 


12 400 


SO 200 


36 200 




Ravenswood grano-diorite 



9 800 




6 000 


18 400 


8e 8nn 

8,1 600 

8l 200 

Summary of quality 















Good rock, 1st grade.. . 

. . 42 400 




39 800 


Probably fair, 2d grade. 

... 30 800 


34 800 

(41. I) 

13 600 


Probably poor, 3d grade. . 

. . 6 600 



(II. 8) 

9 400 

(11. 6) 

. . 6 000 


6 000 


18 400 






8l 200 


Argument on choice of line 
In judging the quality of rock and its suitability for this con- 
duit the factors of most weight are the same as those repeatedly 
mentioned in connection with other portions of the Catskill aque- 
duct line. That is, in brief, that the harder crystalline rocks of the 
Fordham gneiss 1 and Manhattan schist types wherever known; to be 



free from fault crushing and surficial weathering are the best 
variety ; that the more heavily buried areas of these rocks, together 
with those limestone areas that are known to be the most substan- 
tial of its class, should be regarded as fair or second grade; that 
the more obscure areas of limestone and all portions crossing 
faults or rivers or crush zones in any rock must be regarded as- 
poor or third grade. This rating is based wholly on rock char- 
acter and without any consideration of cost of construction. 

From the above it is clear that line A has more " first grade " 
rock than either B or C and less " third grade " ground. 

Line C has three times as much " unknown " ground as either 
B or C and less " first " and " second grade " rock. 

In other words, the three lines are estimated : 

First grade rock 

Second grade rock 

First and second grades together 

Third grade rock 

Unknown ground 




Per cent 

Per cent 

Per cent 





41. I 






11. 8 

11. 6 




In addition to these differences of quality, it appears from a 
study of the areal geology along the respective lines that a tunnel 
would pass across limestone contacts from one formation to an- 
other six times on line A, four times on line B, and seven times 
on line C. These may all be considered points of probable 

All of the lines cross belts of well known weakness believed to 
represent fault zones. Line A crosses three such zones, line B 
crosses two, and line C crosses at least three. 

Furthermore, all of the lines cut limestone for greater distances 
than seems desirable or necessary. The weakest ground and the most 
uncertain quality of ground that can be mapped falls within the 
limestone areas. In this respect line A with 13.9$ of limestone 
ground is preferable to line B, with 25.3$ or line C, with 15.2$. 

From the above it is apparent that line C is least defensible. 
Line A has some advantage over both of the others, especially in 
quantity of first grade rock quantity of first and second grade 
together, low amount of the known poorest grade and small extent 
of the so called " unknown " ground. 

The chief advantage of line A over line B lies in its much 
smaller limestone area (12,000 feet 7'.s\ 21,400 feet or 13.9^ vs. 



25.3$), and the chief advantage of line A over line C lies in its 
much smaller amount of " unknown " ground (6000 feet vs. 18,400 
feet or 7.0^ vs. 22.6$). On these grounds line A is the least ob- 
jectionable of the three lines proposed. 

But it is also clear from an examination of the field, as is shown 
on the accompanying map [pi. 32], that it is possible to avoid 
some of these objectionable features or certain parts of them and 
materially improve the figures by shifting the line to a sort of com- 
promise position between line A and line B. This compromise 
line, or the trial lines from which the final tunnel line may result, 
should follow as closely as possible the gneiss and schist ridges 
and should avoid the limestone areas and known weak zones wher- 
ever possible. 

Depth of tunnel 

The rock formations in general at the required depths are no 
more objectionable on Manhattan island or in The Bronx than at 
other localities on the Southern aqueduct. There are weak places 
and crush zones to be crossed and some of them can not be avoided 
by any possible manipulation of the line, but these most question- 
able spots constitute but a small proportion of the whole distance. 
The depth most suitable must depend chiefly upon the depth neces- 
sary at the worst spots. 

Comparative cost of construction if lines are shifted 

The question is best answered by reference to the geological map. 
It will be noted especially that the belts of the different rock forma- 
tions are usually narrow, and that they run nearly parallel to the 
average direction of the lines. Therefore a shift of line to no great 
distance would at many points place it within an entirely different 
formation. It is also notable that all of the lines run along or near 
the contacts between formations for long distances. At such points 
a very small shift would wholly change the type of rock and rock 
quality. Some shifting is desirable. 

In general it may be assumed that the limestone belts would be 
easiest and cheapest to penetrate wherever they are fairly substan- 
tial, but they undoubtedly also contain the greater proportion of 
weak and troublesome ground and must be considered least desir- 
able from the standpoint of maintenance and durability. The 
gneisses are probably most expensive to penetrate and the schists, 
medium. Both are more expensive than limestone but both arc 
more likely to prove acceptable for other reasons. 



The question of shifting the lines is a complicated one and hinges 
more upon rock conditions, durability, and location of weak zones, 
than on any possible cost. 

Advisable changes in lines 

None of the suggested lines are defensible from a geologic point 
of view for" the reason that a much better one may be obtained by 
no very serious shifting. 

J n the general consideration of relative advantages of different 
possible locations of the line, it is believed that the following large 
features are of most immediate importance: 

1 The ridges as opposed to the valleys. 

2 The hard formations as opposed to the softer ones. 

3 The crossing of few contacts as opposed to crossing many. 

4 The location well within a formation as opposed to location 
along a contact zone. 

It is distinctly preferable from a geologic standpoint ( I ) to fol- 
low the ridges, (2) to keep in the hard formations, (3) to avoid 
many changes from one formation to another, (4) to keep away 
from contact zones, and (5) to avoid weak zones, if possible, or 
cross known troublesome zones at the most advantageous point. 

Recommendations of new lines F, G, H, I 

The original lines A, B and C are marked on the map in blue 
[pi. 32]. In addition several trial lines are sketched in yellow, any 
one of which would give better geological conditions than any of 
the three original lines. The newly suggested trial lines differ from 
each other chiefly in the points at which they cross the limestone 
belts and weak zones. In all of the n the central idea has been to 
follow the gneiss and schist ridges as persistently as possible. All 
unite at Central Park and are intended to follow Fifth avenue, 
Broadway, the Bowery and Market street to Fast river along one of 
the original lines. North of Central Park they differ from the orig- 
inal lines. The westerly one crosses the Harlem river at 176th 
street and may be designated line F. The easterly line may also cross 
the Harlem river at 176th street and may be designated line G; or 
it may continue southward and cross the Harlem at 155th street. 
It will then join the first one in the vicinity of 144th street and is 
called line II. The alternative easterly one which crosses the Har- 
lem at 155th street and follows Seventh avenue to Central Park is 
line I. 



Details of rock conditions along these lines are as follows : 
Line F. (Westerly) beginning at Hill View reservoir 


7 600 Yonkers gneiss — good quality 
15000 Fordham gneiss — good quality 
2 000 Fordham gneiss — probably 2d grade 

1 200 Harlem river crossing — partly limestone — 3d grade 
14800 Manhattan schist — good quality 

1 600 Manhattanville crossing — 3d grade — some limestone 

2 600 Manhattan schist — good rock — through Morningside 


800 At south end of Morningside Park — perhaps some lime- 
stone — 2d grade 
1400 Manhattan sclhst — good — to junction 
12000 Manhattan schist — along Central Park — good 
20600 To East river — Manhattan schist — less known 1 - — -(fair) 
(2d grade) 
6000 To Brooklyn "unknown" 1 

85 600 Line G 


8 400 Yonkers gneiss — good rock 
17 600 Fordham gneiss — good rock 

which brings it to the Harlem river where the other line (F) is 
joined. Although the line is about 1400 feet longer, it avoids some 
low ground (2000 feet) along the east bank of the Harlem river, 
some of which may be in poor condition. Total length of line, 
87,000 feet. 

Line H 


8 400 Yonkers gneiss — good quality 
23 800 Fordham gneiss — good quality — to Harlem river 
1 000 Crossing Harlem river — probably fault zone in gneiss 

800 Fordham gneiss — good quality 
i 000 Limestone — 2d grade 

1 200 Manhattan schist — good quality — to junction with the first 
line (F) at 145th street 

From this point the line is the same as F and G. Its chief ad- 
vantage is the great distance which it has in Fordham gneiss. 
Total length of line, 85,600 feet. 

1 Subsequent explorations made by the Board of Water Supply have climi- . 
nated this unknown ground. See later discussion. 



Line I 


8400 Yonkers gneiss — good quality 
23800 Fordham gneiss — good quality — to Harlem river 

1 000 Crossing Harlem river — probably fault zone in gneiss 
4400 Fordham gneiss — good rock — to 135th street 

4 600 Inwood limestone — probably fair — 2d grade 

2 000 Inwood limestone — probably poor quality — 3d grade 
I 000 Manhattan schist — good quality 

At this point the line unites with line F. Total length of line, 
83,800 feet. 

A tabulation of these figures indicating estimated extent of rock 
types is given below: 

Line f line g line h line i 
Feet Feet Feet Feet 

Total length of line 85600 87000 85600 83800 

Length in Yonkers gneiss 7600 8400 8400 8400 

Length in Fordham gneiss 17000 17600 25600 29200 

Length in Inwood limestone and marginal 

contacts 3600 3600 3400 6600 

Length in Manhattan schist 51 400 51400 42200 33600 

Comparative summary of types of formation (Comparative dis- 
tances are expressed in percentages) 

ABC F G h 1 

Yonkers gneiss 7.2 9.4 7.3 8.8 9.6 9.8 10. o 

Fordham gneiss 8.1 15.3 26.3 19.8 20.2 29.9 34.8 

Contact zones 5.1 0.0 

Inwood limestone 13.9 25.3 

Manhattan schist 58.5 42.6 7.8 60.0 59.0 49.3 40.1 

Ravenswood grano-diorite 1 0.0 0.0 12.0 0.0 0.0 0.0 0.0 

Too little known to classify 1 7.0 7.0 22.6 7.0 6.9 7.0 7.1 

15.2} 4,2 4,1 39 7 8 

1 The Ravenswood granodiorite has been proven by later explorations to 
extend into the territory here marked as too little known to classify. 

As a group it is especially noticeable that the new lines F, G, H, 
I, have a very much lower percentage of contact zones and lime- 
stone. The percentages of gneisses have been notably increased, 
and the unknown and questionable formations have been reduced 
to approximately the lowest terms. 



Estimated summary of quality 


Feet Feet 





Good rock, first grade . . 

53 40O 56 800 


49 600 

Fair second . . 

01 a nr\ t a nr\ 

O ~> 4 OO 

Poor " third " 

2 800 • 2 800 

2 600 

3 000 

I n L' ti ( \w n 1 ( Rrriol<7 1 vn ^ 

V_ 1 1 1\. 1 1 U \\ 1 1 \ J_> 1 UUMJ 11 i . . 

fi 000 fs noo 

6 000 

6 000 

85 600 87 000 

85 600 


1 All of this rock is now 

known to be of good quality. 

In other words these 

new lines show : 


Per cent Per cent 


Per cent 

Per cent 

• • 62.3 65.3 


59- 1 


27.3 24 


26. 1 




89. i 

3-2 3 



" Unknown " ground 1 .... 

7.0 6 




1 Results of recent boring explorations show that this ground is first 
grade also. 

A comparison on this basis with the original lines A, B, C indi- 
cates that these new lines F, G, H, I, make a better showing, 
especially on first grade rock and that all show decided reduction in 
the third grade ground. 

ABC F G h 1 

First grade rock 49.4 40.0 49.0 62.3 65.3 63.8 59.1 

Second grade rock 35.9 41. 1 16.7 27.3 24.6 26.1 30.0 

First and second 85.3 81.0 65.7 89.6 89.9 89.9 89.1 

Third grade rock 7.7 11. 8 11. 6 3.2 3.0 3.0 3.6 

Unknown 1 7.0 7.1 22. 7 7.0 6.9 7.0 7.1 

1 Now known to be first grade. 

On geological grounds, therefore, it is confidently believed that 
any one of the new lines (F, G, H, I) would give decidedly better 
results than any one of the original ones (A, B, C). The poor 
and the questionable and the unknown ground can not be wholly 
avoided by any possible line, no matter how roundabout, in these 
lines, approximately as drawn, the objectionable points are reduced 
to a minimum with almost no increase in total length of conduit. 
The objectionable portions are also restricted in large part to the 



Harlem river, where we already have the experience of the last 
aqueduct (the New Croton aqueduct) as a guide, and a very few 
other spots. 

General conclusions 

Line I is the shortest possible defensible line. Its chief objec- 
tionable feature is a rather long stretch, 6600 feet of limestone, 
from 135th street to Central Park, upon the quality of which there 
are no data. It crosses the Harlem river fault probably in gneiss. 
But it crosses the extension of the Manhattanville fault in lime- 

Lines F, G and H are almost equally defensible. Line G is 
longest, but is in some respects — especially in following the ridge 
crests — one of the best possible locations. 

It should be appreciated that many other matters, such as 
municipal works already completed or projected, or matters of 
engineering practice, are likely to make it necessary to modify any 
line proposed, and that the final line is more likely to be a com- 
promise, considering all interests. 

A graphic representation of the comparative merits of the pro- 
posed lines is given in plate 33. This is strictly a geologic study. 
The lines are properly placed on an outline map of the city corre- 
sponding exactly to those drawn on the geologic map, plate 32. 
The geologic formations that each would cut are represented on 
longitudinal sections which follow each line, and the attitude and 
structure of each formation are indicated. 

Revised lines 

Subsequently two revised lines based upon the preceding studies 
were examined to determine preference. Later one of these, or a 
slight modification of it, was adopted as the one to be explored. 
It was soon determined on the same reasoning as was applied to 
the first group of lines that the most westerly line — the line keep- 
ing as much as possible within the gneiss and schist ridges — 
would be the most likely to give satisfactory conditions. By this 
method of selection the unknown or untested and doubtful ground 
was reduced to its lowest limits. It was found that nearly all of 
the very weak spots could be located by inspection in the northern 
portion of the line, but south of 59th street the question is de- 
cidedly more difficult because of the heavy drift cover. No rock 
outcrops occur south of 30th street, and one is reduced to the 
evidence of deep borings. 

Geologic detail of the Manhattanville-Morningside section showing the alternative lines studied, the locations of exploratory borings, the two 

principal crush zones and longitudinal profiles 



Points for exploration north of 59th street 

It was soon evident that extensive exploratory work would have 
to be undertaken and the following points were selected at which to 

1 The Harlem river crossing, where the distribution conduit line 
crosses the river just below High Bridge [see later description]. 
The only good evidence as to character of rock at this place is from 
the pressure tunnel of the New Croton aqueduct which crosses the 
river a short distance above. 

2 The Manhattanville cross valley (125th street depression). 
This is the most important cross depression on the island of Man- 
hattan. It is apparent after a little investigation that the bed rock 
floor lies deep and that if it were not for the drift filling the tides 
would surge through this valley making a direct connection between 
the Hudson and East river. It was the least known as to depth 
and character of any point along the proposed line. 

3 The depression between Morningside and Central Park. At 
that place limestone on the crest of a pitching anticline reaches 
farther south than on either side and is more deeply eroded. The 
other zones of large importance are in southern Manhattan the 
geology of which is a special study. 




The necessity for exploration in certain sections of this area can 
not be appreciated without a statement of the local geology and 
especially of the revision of both areal and structural geology that 
the writer has based upon an exhaustive study of all the available 
drill cores and other data to be found in southern Manhattan, East 
river and Brooklyn. 

Below Central Park there is now little geology to be gathered 
from a study of the present surface. But as far south as 31st 
street the bed rock geology is pretty well known from earlier re- 
ports and from recent improvements that have exposed the under- 
lying rock. All of this portion is mapped as Manhattan schist ex- 
cept one small area of serpentine at 59th street between 10th and 
nth avenues. There is no reason to modify this usage. A careful 
study of a great number of rock borings from the Pennsylvania 
Railroad tunnel across Manhattan at 32d street proves beyond 
question that bed rock is Manhattan schist, including almost all 
known variations and accompaniments, for the whole width of the 
island along that line. 

Still farther southward the points that have yielded exact in- 
formation about bed rock are less numerous, and below 14th street 
are confined to deep borings or an occasional very deep excavation 
for foundations. Even these sources of information are lacking 
over large areas. The greater number of borings available are 
along the water front. Their distribution is such as to indicate that 
the west side and central portion and southerly extremity of the 
island are all underlain by Manhattan schist. This is true eastward 
to the East river at 27th street, and as far eastward as Tompkins 
square at 10th street and almost to the Manhattan tower of Brook- 
lyn bridge in that vicinity. 

To the eastward of these limits, i. e. to the eastward of the line 
projected from Blackwell's Island to the Manhattan tower of 
Brooklyn bridge, there is a more complicated geology. The borings 
of the East river water front are decidedly variable. They are 
certainly not all Manhattan schist of the usual types. Those most 
unlike the Manhattan are at the same time most like some varieties 




of the Fordham, and indicate that these formations both occur. 
The lack of any data in the beginning of this investigation except 
on the water front made it impossible to draw more than very gen- 
eral lines. Drawn in this way, the lines of course are too straight, 
but it is certain that they indicate more nearly the actual existing 
areal distribution of formations than any of the maps now in exist- 
ence. 1 They indicate a southward extension of the Blackwell's 
Island belt of Fordham gneiss toward the Manhattan tower of 
Brooklyn bridge. How much of this anticlinal fold of Fordham 
actually brings this formation to the surface it is impossible to say, 
but that it may be expected to be encountered along this line is 

On the east side a parallel belt of Inwood limestone is indi- 
cated and this again is succeeded by a Fordham gneiss area 
which occupies the rest of the eastern margin. Explorations 
made along the line of the gas tunnel across East river at 72d 
street 2 indicates comparatively narrow belts of limestone there 
in both the east and west channels. The limited width of limestone 
at these points, together with the occurrence of two strongly de- 
veloped disintegration zones, seem to indicate rather extensive 
squeezing out and faulting of this formation along fault planes 

1 In the summer of 1908 the writer was assigned the task of studying 
in detail the evidences of geologic structure beneath the drift in southern 
Manhattan. Before any drilling was attempted in the city by the Board 
of Water Supply, a thorough canvass was made of all previous borings in 
this district and the cores and records were personally inspected. More 
than 300 such borings were found in which some of the core could be 
secured for identification and classification as to formation and condition. 
Most borings were given no weight at all in the final summary of this 
evidence unless the rock core or at least fragments of it could be secured. 
After all of these newly assembled data were tabulated and plotted on the 
map, it was evident that if the identifications were correct the areal and 
structural map of southern Manhattan needed extensive revision. A new 
map therefore was made and presented to the chief engineer of the Board, 
October 30, 1908. This has been used since as the basis for exploration of 
the Lower East Side section. This original tabulation and map only 
slightly modified was published under the Areal and Structural Geology of 
Southern Manhattan Island [N. Y. Acad. Sci. Annals, April 1910, v. 19, no. 
11, pt 2]. The extensive explorations of the board have made further revision 
necessary [see accompanying map, pi. 34]. Exploratory boring is still in 
progress (October 1910) and some slight modifications of boundary lines 
may yet be made. 

2 This is taken from Prof. J. F. Kemp's description of The Geologic Sec- 
tion of the East River at Seventieth Street, New York [N. Y. Acad. Sci. 
Trans. 1895. 14:273-76]. 

N. Y. State Museum Bulletin 146 

Plate 34 

e map reprtfdurtHl frt 

5 Beekman street, and ber* used by permlssl 

Margin- of Long Island 

This revision 

Revised Areal Geology op Southern Manhattan Island and the Adjacent 
is based upon exploratory borings to June 25th. 1910. The heavy blue line marks the course of the proposed pressure tunnel intended to carry 

the Catskill water to Brooklyn 



parallel to the strike. Such movements are capable of cutting out 
the intermediate limestone entirely from between the schist and 
gneiss. How much of such modification exists, in the almost total 
lack of data bearing upon the question, it is impossible to say. The 
intermediate belt is indicated on the accompanying map [pi. 34], as a 
limestone area. At one point at least the limestone does occur in 
the older borings, i. e. on the southeastern margin of the Man- 
hattan pier of the Manhattan bridge (bridge no. 3), at the foot of 
Pike street. 

On the Brooklyn side no formations of this series except the 
Fordham and its associated igneous masses, such as the Ravens- 
wood granodiorite, have been identified within the area under study. 
Limestone is reported (Hobbs reference to Veatch) near Newtown 
creek, a little beyond the eastern margin of the present map. 

Structure of the East river area 
Manhattan side. In all of the area south of 59th street, 
structural features are even more obscure than the areal geology. 

There is no reasonable doubt but that weak zones will be found 
as frequently in the Manhattan schist portion of this area as on the 
line north of 59th street, but they can not be indicated as closely. 
No cross fault of large consequence can be identified, but there is 
some evidence of a minor zone that should be encountered on Fifth 
avenue, in the vicinity of 32d street. The Pennsylvania tunnels and 
the subway both cross this line and so far as known there were no 
serious weaknesses developed. There is nowhere any evidence of 
an important depression like the Manhattanville valley. 

It is confidently believed that the problems on this southerly por- 
tion of Manhattan are involved chiefly with the longitudinal struc- 
tures produced by folding and faulting and subsequent disintegration 
along such zones. 

Crossing of East river 

From 59th street to the East river there seems to be no reason 
for a preference between the two lines P and Q. 1 On the Brooklyn 
side likewise there is no known geological reason for preference. 
Such basis for choice as is now known relates to the East river 
channel alone. Since this is at the same time the most difficult sec- 
tion of the line to explore and probably the most uncertain section 
to estimate as to condition and consequent depth of tunnel, it would 
be especially useful to be able to make a decisive selection of cross- 
ings at once. 

1 For location of these lines sec map, pi. 32. 



Such evidence as has any bearing upon this question has already 
been used in formulating the interpretation of geologic structure 
given in the foregoing sections of this report. If the succession and 
boundaries of formations as outlined are reasonably close to the 
actual conditions, it would appear that line P (the southerly one 
just above Manhattan bridge ) lias some advantage over line Q 
(near Williamsburg bridge). The chief elements in this advantage 
are as follows : 

1 It would appear that line P might lie wholly within the Ford- 
ham gneiss in the East river section, while line Q may cross two 

2 From the evidence of borings made in the East river at 14th 
street 1 it appears probable that a belt of schist similar to Manhattan 
schist in quality (whether accompanied by limestone or not there is 
no direct evidence) lies in the river channel toward the east side and 
in all probability extends southward in the middle of the river at 
Williamsburg bridge. This would be cut by line Q. The uncertain- 
ties of this association are of sufficient importance to throw the bal- 
ance of present choice toward line P. 

3 If the theory that the East river course is due chiefly to zones 
of weakness following fractures or faults is true, their possible 
comparative condition as they cut through different formations must 
be taken into account. There is little doubt on this point but that, 
in zones of similar original disturbance, those in the Fordham 
gneiss have suffered less extensively from disintegration than those 
cutting either the limestone or schist. Therefore, obscure as it may 
be, the preference is again in favor of line P. 

4 If, furthermore, the course of the river is due to cross faulting 
or any similar or related displacements or movements, an inspection 
of the structural map indicates that the controlling zone followed 
by the river as it crosses line O must have a general strike north- 
west, while the corresponding zone that crosses line P strikes east. 
Of these two types (directions) of fault zones, so far as they may 
be judged to have influence in the adjacent area, there is no doubt 
but that the northwest type (the set that has a northwest strike) 
is both the more common and the more important. If this general 
tendency is also true here, then on this account also line P may be 
considered slightly more favored. In reality not much weight can 

1 These borings were made by the Public Service Commission in explora- 
tions for subways. 


be given to this point since the condition of these faults is not fully 

5 If, as may well happen, the present East river is displaced 1 from 
its old channel by glacial drift, so that it is essentially an evicted 
stream, there may not be as pronounced a channel or as weak ground 
to cross at such points as at those where the old channel is still oc- 
cupied. In such case both of these lines are favorable. 

6 On the other hand, the crossing of line P is almost a mile 
nearer to the great Hudson gorges, to which doubtless this portion 
of the preglacial East river was tributary, and consequently its bed 
rock channel, if it is the real preglacial channel, may be expected 
to be deeper and the accompanying disintegration (so far as it may 
be controlled by this factor) may be expected to reach lower than at 
points in similar surroundings farther up stream. It is impossible 
to say how much weight should be given to this objection. It does 
not seem to be of sufficient importance to fully offset the favorable 
features indicated in items 1, 2, 3 and 4. 

On the basis of these studies line P (the southerly one) near 
Manhattan bridge was chosen as the site of preliminary exploration 
promising the most favorable results. Eater this was shifted a short 
distance without introducing any new conditions. 

1 Exploratory borings indicate that such has been the history of the river. 



Exploration by borings 1 and other methods have been made at all 
questionable or uncertain points along the line. As was expected 
in the beginning five places have required elaborate exploration and 
some exceptional conditions have been proven. The original 
geological investigation based upon surface study as outlined in the 
foregoing pages served to locate these spots accurately. 

These places or zones, now either finished or sufficiently well 
known to permit accurate statement of geologic conditions, are as 
follows : 

1 The Harlem river crossing at 167th street, where the aqueduct 
will cross from a ridge of Fordham gneiss beneath the Harlem river, 
where the whole thickness of Inwood limestone will be cut, to the 
ridge of Manhattan schist above the Speedway on Manhattan island. 

2 The Manhattanville cross valley, a low pass crossing the island 
at about 125th street. The part explored extends from St Nicholas 
to Morningside Parks and crosses a zone with very low rock floor 
in the Manhattan schist. 

3 From Morningside to Central Parks. The line crosses the 
strike of the formations at this point and cuts a longitudinal fault 
and anticlinal fold which tends to bring the Inwood limestone 
within surface influence. 

1 Exploratory work has been in direct charge of Mr T. C. Atwood, division 
engineer, who has followed all stages of it almost from the beginning. In 
the later exploratory work an immense amount of detail and a very com- 
plex lot of data has accumulated requiring constantly the services of a man 
with some special geological training. Mr John R. Healey, formerly in the 
testing laboratory, was transferred to this special field. He is probably 
more familiar with the multitude of details resulting from boring operations 
along the conduit line than any one else. Except for the care and good 
judgment used by these men in preserving data, and the wisdom of the 
men who planned the line and methods of work before them, much valu- 
able geologic data would have been lost. Notwithstanding the best efforts 
of the consulting geologist some really critical points escape unless some one 
constantly on the ground is directly interested in them as a part of the 
regular responsibility. 


2 3 8 


4 The Lower East Side zone. On Delancey street east of the 
Bowery, the line crosses the structure and at this point the whole 
series of crystalline formations appears. Besides complicated struc- 
ture there is also exceptionally deep alternation or decay of bed rock. 

5 The East river crossing — from the foot of Clinton street to 
Bridge street, Brooklyn. 

i Harlem river crossing 

Geologically the Harlem river between 155th and 200th streets 
has the same relation to local formations for the whole distance. 
It flows on the Inwood limestone bed which stands almost exactly on 
edge, while the east river-bluff is formed by the underlying Fordham 
gneiss, and the west, by a strong escarpment of Manhattan schist 
which extends southward throughout the whole of Manhattan form- 
ing the backbone of the island. 

At the selected crossing a short distance below High Bridge, near 
167th street, the schist-limestone contact is in the river and appears 
to be a low weak spot [see detail of record]. The limestone-gneiss 
contact however is in the flat east of the river bank, near Sedgwick 
avenue and seems to be more substantial. The structural detail and 
relations are shown on the accompanying profile and cross section, 
[Pi. 35]. 

It is observed by examination of the data secured by borings that 
the limestone formation at this point is exceptionally heavily im- 
pregnated with pegmatite dikes and stringers, and that interbedded 
schist layers are large and numerous. 

The weakest spot found lies at the contact between schist and 
limestone where there is probably some longitudinal displacement. 

A similar condition was found at the new Croton crossing 2000 
feet farther north. On the whole bad decay does not extend very 
deep — 150-200 feet. 

Several borings have been made and on them is based the only 
judgment possible of the actual structure and physical condition of 
rock. In most cases the evidence is easily interpreted for these 
points. The most weakened spot, as well as the most difficult to 
interpret in all its detail, is the limestone-schist contact. It is 
judged that hole no. 17 cut through this contact zone. This boring 
is located in the river 50 feet from the Speedway (west bank) on 
the proposed tunnel line which crosses a short distance south of 
High Bridge. It is known as hole no. 17/C38. Because of the 


somewhat unusual quality of material at this place as indicated by 
the wash and core saved and because of the suggestion it gives 

Fig. 38 Key map showing plan of exploratory borings at the Harlem 
river crossing, location of the New Croton aqueduct which crosses the 
Harlem in a pressure tunnel and the Old Croton aqueduct which crosses the 
river on High Bridge 

about the structure and condition of rock beneath the river, the 
record and interpretation notes are given. 




o — i3=Water 

13 — 46=Black river mud (mostly river silt) 
46 — 48=Sand with decayed wood (peaty wood) 
48 — 7o=Quartz and garnet sand rather clean (glacial) 
46 — 70=Lumps of peaty matter coming to the surface at inter- 
vals indicating occasional small layers of peat 
70 — 78=Mixed sand (glacial) 

9 2 =A core of Triassic contact shale (a drift boulder from 
the Palisade margin). At this point also a piece of 
Manhattan schist (boulder) 

95 =4 pieces of diabase (Palisade trap) from another drift 

9^-5 =5 pieces of Inwood limestone (boulder) followed by a 
piece of quartzite and several mixed pebbles indi- 
cating glacial drift origin 

114 — II9=A buff yellow sand with much pearly yellow mica flakes. 

Effervesces with acid. This shows no foreign matter. 
It is chiefly residuary decayed rock in place and repre- 
sents silicious and micaceous limestone. It is decayed, 
very impure, Inwood limestone 

119 =Clay with pieces of flinty quartzite, probably from a 
small quartzose seam in the limestone 

120 — i2(5=Light flaky yellow material. Much pearly mica with 
earthy matter. Effervesces in acid. Residuary from 
Inwood limestone 

128 =White and drab lumpy residuary matter (kaolin) and 
earthy substances. Effervesces. A more impure In- 
wood. Also shows several pieces of core of a porous, 
rotten limestone. Inwood 

129 — i34=Reddish brown lumps. Effervesces a very little. 

Mostly clay but still no foreign matter. Residuary 
material from a more silicious bed. A few pieces of 
hard, impure limestone at 133 feet 

134 =Pieces of a porous quartz chlorite rock with' little lime. 

Is a leached quartzose rock evidently a sandstone 
layer in the limestone. Rock belongs to the Inwood 




135 — i43=Dark micaceous matter containing chiefly biotite, a 

pearly mica, and quartz. Rock is a decayed schist 

bed=the transition between Inwood limestone and 

Manhattan schist 
143 — i5i=Dark brown micaceous material. Biotite and quartz — 

chiefly. Rock is decayed schist (transition rock). 

At 146 feet encountered pieces of a pegmatite veinlet. 

All pieces except 1 are pegmatitic — the other one is 

calcareous sandstone, fallen into this lot from the 

134 foot level 
151 — i6o=Chunks of pegmatite (a vein rock) 
151 — i6i=The mica washings continue the same as at 143 — 151 

feet. Rock is a transition schist with pegmatite 


164 — i6o==Brownish yellow micaceous matter (loose). Mica, 
quartz, chlorite, lime. Effervesces 

164 — i73=Many pieces of typical Manhattan schist. A fair 
amount of core for the conditions. Rock is not so 
badly decayed but is broken into small pieces. Rock 
is Manhattan schist of typical character. 


1 The material is chiefly river silt down to 46 feet 

2 Lighter glacial deposits 46 — 78 feet 

3 Heavy bouldery drift 78 — 97 feet 

4 Uncertain (insufficient data) 97 — 114 feet 

5 Residuary micaceous decay products from Inwood limestone 

114— 135 feet 

6 Decayed transition schist bed with some lime, but chiefly like the 

Manhattan schist 135 — 161 feet 

7 More calcareous schist 161 — 164 feet 

8 Typical Manhattan schist 164 — 173 feet 


1 Foreign matters, glacial and recent deposits, continue to a depth 

of between 97 and 114 feet. 

2 Rotten formations (residuary matter) in place begin at least as 

high as 114 feet. There is no foreign material below that point 
except grains that have fallen into the hole from above. 



3 More solid rock begins at 164 feet. 

4 The upper portion of the rotten rock (114-35 feet) is calcareous 

enough to belong to the Inwood limestone formation. The 
lower 9 feet (164-73 fe et) is typical Manhattan schist. The 
intermediate ground 135-64 feet is transition variety. 

5 The drill has cut the contact between Inwood and Manhattan 


6 If this identification of the badly decayed matter is correct, the 

contact at this point dips steeply eastward, i. e. it is overturned. 

7 Both types of rock are shown to be extensively decayed. 

8 The worst (deepest) decay zone probably lies still a little farther 

east, and follows the dip of the micaceous limestone near the 

These conditions are indicated on the accompanying cross section 
[see pi. 35]. 

The conditions indicated by this one hole are consistent with 
those known for the New Croton aqueduct tunnel 2000 feet farther 
north where, according to the engineers' drawings, the formations 
also are overturned. Fifty feet of decayed rock is shown in this 
hole. The contact is undoubtedly decayed considerably to a depth 
of more than 200 feet below water level. 

Fig. 39 Harlem river crossing — New Croton aqueduct 

Another boring put down to test conditions at still greater depth 
nearby explored the rock to -442.7 feet. Because of the informa- 
tion it gives about the deeper bed rock, a summary of the record 
based upon examination of the material is given : 

Geologic cross section and graphic interpretation of the exploratory borings made for the New York City Board of Water Supply at the site of the proposed pressure tunnel beneath the Harlem 

river, reaching Manhattan at the foot of 171st street 


- IC 





Hole no. 42 (75 feet from Speedway, 25 feet east of hole no. 17) 




River muds and various types of drift similar to 

hole no. 17 




Iron cemented sand — both drift sand and local 

angular material 




Micaceous clay — residuary decayed matter — 

with choppings of calcite, quartz, mica and 

chlorite representing weathered Inwood lime- 





Core — Inwood limestone (impure) 








Inwood limestone — typical — standing almost 

vertical in upper portion but changing to about 

45 and farther down to 60°. Good sound 





Manhattan schist — of typical sort — and in 

sound condition, but becoming somewhat more 

broken and altered near the bottom. Dip about 

6o°-8o° and even more. Average probably 





Manhattan schist — typical — dip variable but 

mostly above 70 to vertical — some pegmatite 

— fractures are at high angle. Rock sound 








Inwood limestone — typical — good quality — dip 

70 to very flat — one piece not over 35 but 
mostly obscure 
An interpretation summary is as follows: 


o to -94 River muds and drift filling (glacial and recent) 
-94 to -96 Transition to residuary matter 
-96 to -127 Residuary matter and badly decayed Inwood lime- 

-127 to -197 Inwood limestone 
-197 to -302 Manhattan schist 
-302 to -442.7 Inwood limestone 

Geologic cross section. The accompanying cross section 
[pl- 35] embodies an interpretation of all the data secured in the 
Harlem river. It is now known that the limestone is overturned 



slightly at both contacts. The nature of these contacts makes it 
seem probable that there is very little of the limestone squeezed or 
cut out by movement. Therefore this crossing gives a fairly ac- 
curate measure of the thickness of the Inwood. This is approxi- 
mately 750 feet. No section about New York city is more accurately 

2 Manhattanville cross valley 

In northern Manhattan the schist ridge which forms the back- 
bone of the island and has a relief of more than 100 feet, is cut 
across by a prominent valley that extends from the Hudson at 
130th street eastward to the Harlem Flats and East river. This 
valley is nowhere more than 25 or 30 feet above the sea level and 
is drift filled. Previous to the recent boring explorations of the 
Board of Water Supply its true depth to rock floor was unknown. 
The few borings recorded, however, indicated a depth of more than 
a hundred feet. One such boring at 129th street and Amsterdam 
avenue is reported as penetrating 109 feet from surface without 
touching rock. Another of similar results is located at 125th and 
Manhattan streets where a depth of 204 feet failed to touch rock. 

Besides determining rock floor in the present case, it was im- 
portant to determine rock structure and conditions. It appears 
from surface features that this cross valley probably follows a 
fault zone along which there has been weakening of the rock and 
consequent disintegration and decay. If this is so it would be ad- 
vantageous to find the limits of it and determine what displace- 
ment effects were produced. It has been surmised by all students 
of local geology that such cross faults may lift the blocks on the 
south side of them, one of the chief indications being the fact 
that in spite of a strong southerly pitch in all the formations they 
do not rapidly disappear below sea level. 

The accompanying profile and explanatory section indicates the 
principal results of exploration [see pi. 36]. Badly crushed 
ground has been found in the holes near the north end of Morning- 
side. Park but the rock, when found, is not very badly decayed. 
The rock floor is very low, almost 200 feet below sea level at the 
lowest. It appears that if the drift were stripped off from this 
valley the Hudson and Long Island sound would unite across the 
Harlem Flats and Manhattanville forming a channel and outlet 
much deeper than the present East river course. 

The glacial drift of this valley is prevailingly fine modified drift 
some of which is probably stratified and fairly well assorted. 



This is more strikingly true of the southerly extension of this low 
ground southward along Morningside Park. A very deep and 
prominent preglacial stream came down from the gap between 
Morningside and Central Parks. 

It is not yet proven that the fault has really raised the Morning- 
side block. At least if there is such displacement it is not of suf- 
ficient amount to bring up a different formation at any point- 
yet examined. It would be possible for the limestone to be brought 
up to the surface, but except for a few pieces of interbedded lime- 
stone no evidence has been secured. The occurrence of this, how- 
ever, is thought to indicate proximity to the limestone contact. 

General geologic conditions established. Fourteen borings 
have been made for the special purpose of determining exact con- 
ditions. On the data of these holes there are several features now 
established beyond question that were originally given only as prob- 
abilities. The most important of these may be enumerated as fol- 
lows : 

1 A very deep cross valley is now proven between 123d and 
126th streets, and its profile can be plotted. 

2 A part of this ground is badly broken, as if belonging to a 
fault zone, but most of the floor thus far tested is not in had con- 
dition, i. e. it is not very badly crushed or decayed. 

3 The drift cover in this cross valley is more than 200 feet deep 
over a distance of more than two blocks on the proposed line (from 
123d street to Manhattan street). 

4 The limestone contact lies more than 300 feet east of the pro- 
posed line at this Manhattanville cross valley. 

5 At 121st street the limestone-schist contact stands very steep 
and is probably slightly overturned. This is indicated by the data 
of hole no. 33. 

6 The contact line approaches nearer to Morningside Park in 
passing southward, touching the park between 110th and 113th 
streets and the contact is probably not overturned in this southerly 

3 Morningside to Central Parks 

The contact between In wood limestone and Manhattan schist 
follows nearly parallel with the Morningside Park boundary on 
the east side, but, because of its form, actually touches the park 
only at the southern end between noth and 113th streets. At the 
north end it lies off more than half a block to the east. The Man- 
hattan schist forms an escarpment because of its more resistant 



character and this eastward facing cliff and slope forms Morning- 
side Park. St Nicholas Park, farther north, from 128th to 155th 
streets has the same structural relations. In both cases the present 
escarpment stands back from 200 to 500 feet from the actual 

As the formations all pitch southward and are pretty closely- 
folded, the higher formations gradually appear and at 110th street 
another parallel ridge of Manhattan comes in above the limestone 
in the trough of the next syncline to the east. This forms the 
north end of Central Park and from this point southward Man- 
hattan schist is continuous. But between the Morningside belt of 
schist and the Central Park belt at 110th street lies an anticline 
of Inwood limestone also pitching southward and gradually pass- 
ing beneath the schist which encroaches upon it in a long wedge 
until a few blocks farther south it passes wholly beneath the schist, 
which from that point is continuous. 

This anticlinal wedge and its accompanying structures and rock 
condition was the subject of some detailed exploration. 

The records of a few drill holes together with an interpretation 
of all the data will serve for the present purpose. 

The most important borings are summarized below : 

a Hole no. 3 on 113th street, 232 feet east of Morningside Park 

Surface elevation-l-42.6 feet 
Rock floor at depth of 81.5 feet=el. -38.9 feet. 
Material : 

0-19 feet=to eL+23.6 feet=soil and mixed drift 
19-79 feet=to el. -36.9 feet=modified drift. Assorted sands 
and silts 

81.5-94.58 feet=to el. -54 feet = Inwood limestone. Typical 
and in good condition 
b Hole no. 7. On 113th street, corner of Manhattan avenue 
Surface elevation+38 feet 

Rock floor at depth of approximately 165 feet=el. -127 feet 
Material : 

0-85 feet=to el. -47 feet=modified drift 

85-165 feet=to el. -127 feet=sand with much more clay, part 

of which may be decayed rock 
165-240 feet=to el. -202 feet=disintegrated rock ledge. Some 

micaceous type believed to be the transitional facies of the 

schist-limestone contact 

N. Y. State Museum Bulletin 146 

Plate 36 


t-J~i L_J ISIi Lsjj i_iL t_Jio 1 , .AVE, ^ ^ i_ 

"1 □ n (Tj m ra R]i^ r 
^gS^rm'ra 41 n n i i r 

"T^Ti^ K ^ # i g 33 If- 


uy •■■ 1 m i t if 1 1 [ 
-j MOBMN6S10E p* g£ 

n 3 ! I — 1 r~p] 1 — 1 f-pi r" 


Geologic detail of the Manhattanville-Morningside section showing the alternative lines studied, the locations of exploratory borings, 

principal crusR zones and longitudinal profiles 



242-280.71 feet=-to el. -242.71 feet=Inwood, very coarse type 
of limestone. Poor core showing. Muoh broken 
c Hole no. 12. In Morningside Park at 113th street 
Surface elevation+28 feet 
Rock floor at depth of 84 feet=el. -50 feet 
Material : 

0-26 feet=to el.+2 feet=mixed drift 
26-84 feet=to el.— 56 feet=modified drift 

84-335.15 feet=to el.-307.15 feet= Manhattan schist, typical 
with considerable pegmatite. But all good sound rock, not 
much broken and standing at about 65°-8o G 
d Hole no. 16. Corner of Manhattan avenue and 110th street 
Surface elevation=+55 feet 
Rock floor at depth of 159 feet— el-104 feet 
Material : 

0-44 feet=to el.+il feet=filled ground and mixed material 
44-159 feet=to el-104 feet=fine sands and silts interpreted 
as chiefly modified drift. Much of it very fine and the lower 
portion rather micaceous and angular throwing a little doubt 
on the exact line of demarcation between drift and residuary 

159-161 feet=to el-106 feet=core of Manhattan schist 

1 71-186 feet=to el.— 131 feet=decayed rock in place, some 
micaceous type, coming out as mud 

186-228 feet=to el.— 1 73 feet=micaceous reddish mud with 
variable amounts of angular quartz grains. Certainly residu- 
ary decayed rock 

228-270 feet=to el-215 feet=similar residuary matter less 
highly colored passing from reds into grays and coming out 
as soft material 

270-305 feet=to el-250 feet=grayish micaceous and quartz- 
ose residuary matters. With much silvery mica and chloritic 
grains near the bottom 

305-335 feet=;to el-280 feet=Inwood limestone, core, ordi- 
nary type. No more recovered above this point except for 2 
feet between 159 and 161 feet 
c Hole no. 36 at 108th street and Manhattan avenue 

Elevation of surface+63 feet 

Rock floor (decayed) at depth of 55 feet— el .+8 feet 
Depth to solid core=248 feet^el-185 feet 



Material : 

0-55 feet=el.+8 feet=modified drift (fine silts) 

55 _1 55 feet=:to-io8 feet=micaceous soft material with 
broken sand=decayed micaceous rock 

1 55—2 1 5 feet=to-i52 feet=reddish mud of similar constitu- 
ents. Is decayed rock colored by iron 

215-240 feet=to-i 77 feet=transition to ir.ore grayish and 
greenish soft matter 

240-245 feet=to-i82 feet=greenish mica rock=a decayed 
chlorite, mica quartz, schist layer 

248.33-254.25 feet=from el.-185.35 to-191.25 feet=chloritic 

Inwood limestone 

A summary of these data gives : 
0-55 feet=drift 

55-245 feet = decayed rock ledge 
248-254 feet=solid rock ledge (limestone) 
/ Hole no. 2 at 123d street, 100 feet east of Morningside Park East 
Surface elevation+30 feet 
Rock floor at depth of 220 feet=el.-i90 feet 
Material : 

0-13 feet=to eL+17 feet=soil and mixed drift 

13-220 feet=to el-190 feet=modified drift=mostly assorted 

sands and silts 
220-245 feet=to eL-215 feet=soft decayed schist 
245-355 feet=to el-325 feet=Manhattan schist much broken 

— poor core recovery — worst material at about 225-240 feet 

and again near bottom. Formation evidently much shattered 

and considerably decayed 
g Hole no. 33 on 121st street, 300 feet east of Morningside Park 


Surface elevation+31 feet 
Rock floor at depth of 195 feet=el.-i64 feet 
Material : 
0-25 feet — soil and mixed drift 

25-195 feet=to eL-164 feet— drift, mostly modified drift= 

assorted sands and fine silts 
190-195 feet coarser material — pebbles 

195-200 feet=to el-169 feet— Inwood limestone, coarser lime- 
stone of usual type 

200-237 feet=to el-206 feet=Manhattan schist 

Ordinary type and in good condition [for interpretation see 
later comments] 



Condition of the limestone schist contact. The finding of 
Inwood limestone above the Manhattan schist in hole no. 33 at 
121st street east of Morningside and the fairly sound condition of 
both typeis raises the general question of the condition of contact 
zones as compared with fault zones. 

There are three important facts to consider bearing on this case : 
(1) The contact zones are commonly weaker than either formation 
alone and (2) at this particular point an abnormal relationship is 
shown by the overturned strata (the limestone lying above), and 
(3) the fault zones are always weak and extensively decayed. 

Because of the abnormal position of the limestone here, lying as 
it does overturned, a weaker more pervious rock upon a more sub- 
stantial and less pervious one, it appears to be reasonable enough to 
find the limestone and schist fairly well preserved, under conditions 
where a vertical or a normal position would have encouraged decay 
because permitting a more ready circulation. 

But there is a further conclusion that seems allowable, i. e. the 
fault or crush zones are more extensively decayed than the simple 
contact or transition zones. And contrariwise, where an especially 
extensive decay is encountered, it probably is to be associated with 
a crush zone due to fault movement rather than with any other 

A further inference seems allowable from the data of these holes. 
It is probable that these fault zones do not follow the contacts or 
bedding exactly but cut across at low angles, sometimes coinciding 
with the contact lines and sometimes falling wholly within the lime- 
stone or the schist. 

Great depth of decay at south end of Morningside Park. The 
finding of approximately 150 feet of decayed rock in hole no. 16 
and of nearly 200 feet of similar type in hole no. 36, all so rotten 
that the material came up as a mud, raises a very difficult question 
as to the conditions that make such extensive decay possible. 
Hole no. 7 (113th st.) shows extreme decay to elevation -204 feet 
Hole no. 16 (noth st.) shows similar condition to elevation -250 


Hole no. 36 (108th st.) shows similar condition to elevation -185 

These three holes showing similar condition of very deep decay 
are located almost exactly in line. Nothing on either side of this 
line is in so poor condition. 

Consideration of these conditions can not fail to raise certain 
questions of interpretation. 



1 It would appear that at least one of these borings (no. 7) is 
near the schist-limestone contact. May they all lie then in the 
weakened contact zone? 

2 It is true that at least one core (also from no. 7) shows a 
badly broken condition. May they all lie in a fault zone? 

3 There is no reasonable doubt but that the geologic structure at 
the south end of Morningside Park is that of a pitching anticline 
carrying the limestone beneath the schist in its southward extension. 

May the excessive decay be due to this relation ? 

The evidence on these various possibilities is not complete enough 
to make a conclusion very reliable. But there are two or three 
factors that have a bearing and they unite pretty well in supporting 
one view. 

These factors are: (a) the exact alinement of these three holes, 
(b) the crushed core of hole no. 7. (c) the overturned position of 
the formations 10 blocks farther north, (hole no. 33), together 
with the apparently normal position in hole no. 16. 

All of these points are consistent with the opinion that we have to 
do here with the crush zone of a fault, one that runs rather straight 
and one that follows not far from the contact of the schist and lime- 
stone at this point. And it is probable that the weakness follows 
the west margin or limb of the limestone anticline as it plunges be- 
neath the schist. Such evidence as there is favors this view. 

If that is true, then one may expect that the worst ground is not 
very wide, but that one probably can not go entirely around it. The 
best line would run south far enough to get above the limestone, 
and then cut across the weak zone nearly at right angles. It is cer- 
tain that the ground improves southward. 

Later borings are all confirmatory of the conclusion that the weak- 
ness is narrow and dies out rapidly southward as soon as the lime- 
stone passes well beneath the schist. No bad ground has yet been 
found on 106th street where the tunnel will probably be located. 

4 The East river section 

Preliminary studies of southern Manhattan and the East river 
led originally to the conclusion that the portion of the East river 
forming the great eastward bend from 32c! street to Brooklyn bridge 
probably has a simpler geologic structure than those portions farther 
north or south. It was long known that the structure at Black- 
wells Island is very complex and involves all of the local formations 
in close folding and considerable faulting. But there seemed to the 






writer after studying all available data, good reason to believe that 
the river leaves this belt when it bends to the eastward and that it 
i> in this part a displaced stream. In that case the East river coidd 
bo flowing upon a floor of gneiss of a most substantial sort. 

Explorations are now complete on a line that crosses the river 
from Clinton street, Manhattan, to Bridge street, Brooklyn. All 
borings have found good sound rock at moderate depth and all are 
comparatively shallow holes. Their positions and depths and rock 
types are tabulated below. 

No. of 
bori. g 

Dist nces in feet 
fr-m Manh-ttan 
pier head 

A pproximate 
interval in 

Elevation of 

rock floor 
below mean 
sea level in 

Type cf rock 






2 I 







— 72 





— 71 



1 70 





— 74 





— 81 


1 070 

1 10 




Brooklyn side 

— 75 



near bulkhead 


The rock floor is thus very uniform as to contour across the East 
river at this point. No water course yet explored about Manhattan 
island has shown so simple conditions including as it does sound 
rock and shallow channel. The rock varies a good deal but is pre- 
vailingly a coarse grained granodiorite. In places it is very gar- 
netiferous and at others is banded or micaceous, but all belong to 
the Fordham formation as a general formational unit. 

Borings in the East river made by the Public Service Commission 
both above and below this point found an occasional deep hole with 
excessive decay to more than a hundred feet without securing sound 
core. At this crossing the deepest point in the channel to sound 
rock floor is 81 feet. 

It is certain from these results and from others in adjacent 
ground that the East river does not occupy in this part of its course 
the original stream channel. Tt has been displaced (evicted) by 
glacial encroachment and has never been able to reoccupy the lost 
course. Therefore, instead of the river following a belt of lime- 




stone around this big bend as was formerly supposed, it follows no 
rock floor structure at all but is in this part of its course wholly 
superimposed. The original valley lies farther to the west cutting 
through the midst of the Lower East Side where the more com- 
plicated geologic structures again prevail. 

Borings at intervals of 500 feet have now been made on the 
Brooklyn side of the East river to Gold street and Myrtle avenue. 
So far as developed there is no other formation than the Fordham 
and the associated granodiorite within the area covered. The rock 
floor is remarkably uniform at an elevation of from -70 to -90 
feet. The accompanying section shows the relations of rock floor 
to present drift surface [fig. 40]. 


The proposed distributary conduit turns from the Bowery east- 
ward on Delancey to Allen street, thence on Allen to Hester street, 
thence on Hester to Clinton street and follows south on Clinton 
to the East river. This so called Lower East Side section includes 
one of the most complicated geologic structures in New York city. 
The most complex portion extends from Christie street on the west 
side to Monroe street on the east. Between these two points all of 
the crystalline rock formations form a series of parallel beds that 
are folded together so closely that they stand practically on edge. 

This general fact and the approximate location of the several 
beds have been proven for some time. But the more exact structure, 
with the depths to which the beds go before bending upward again, 
and the distances through each one are only approximately deter- 
mined by the exploratory borings to date. The chief uncertainties 
arise from the fact that the beds are also faulted and the dips of 
the fault planes are not yet determined and the amount of displace- 
ment is unknown. The difficulty of forming a good estimate of the 
obscure points is greatly increased by the fact that no rock of any 
kind is to be seen at the surface. Judgment is based wholly on 

There are other important questions covering the zone, such as : 
(1) depth of serious decay, (2) location and width of these decay 
belts, (3) general physical condition of the rock at certain levels, 
(4) length of tunnel that will cut each formation, (5) best depth 
for safe construction. 

The accompanying geologic cross section [pi. 38] embodies an 
opinion of the structural relations of the different formations. It is 



offered as the writer's interpretation of borings to date, and its 
more direct use is as a working basis and guide in conducting ex- 
plorations. The western half of the section may be accepted as 
more accurate in minor detail than the eastern. 

To simplify the section it is drawn on a line crossing this zone 
more directly than the conduit as laid out, i. e. through holes 28 and 
5 and the borings are projected along the strike of the formation 
to the section line. All the data therefore are used and the structure 
is not distorted, but the distances through each bed would be greater 
on the conduit line because it runs more diagonally across the 

Borings. The following tabulation of borings and interpre- 
tations upon them forms the basis of the present ideas of structure 
and quality of rock on the Lower East Side. The borings are given 
in order from west to east, and all points are neglected except those 
bearing upon geologic structure. 
28 The Bowery and Delancey street 
Surface elevation 40.5 feet 

Rock floor -71 feet. Rock is Manhattan schist, and has been 
penetrated to -91 feet 
78 Delancey street west of Christie 
Surface elevation 41.4 feet 

Rock floor -101.6 feet. Rock all typical Manhattan schist — 
at about 6o° 

224 and 301 North side of Delancey street west of Christie street 

Surface elevation 42 feet 

Rock floor at el. -99 feet 

Manhattan schist to el— 330 feet 

Inwood limestone to bottom at el. -395 feet 
229 Northeast corner Delancey street and Christie street 

'Surface elevation 43 feet 

Rock floor at el. -108 feet 

Manhattan schist with very poor core recovery to el. -260 feet 
Inwood limestone to bottom at el. -360 feet 
84 Delancey street east of Christie 
Surface elevation 41.8 feet 

Rock floor -135 feet. All badly decayed schistose rock, of 
same type — no effervescence — red color — soft as cheese 
to -204 feet 
227 is a reoccupation of this same hole 84 

Inwood limestone was found below el. -250 feet to the bottom 
below el. -300 feet 



63 Delancey street west of Forsyth 
Surface elevation 43.2 feet 

Rock floor —141 feet. Inwood limestone at 80 — very low 
saving of core 
72 Delancey street 121 feet east of Forsyth 
Surface elevation 42.6 feet 

Rock floor -122 feet. Very noncommittal rock, one piece very 
good Fordham and the rest not decidedly any special type 

Classified as Fordham on this basis. Same behavior to bottom 
-109 feet 
81 Delancey street and Eldridge 

Surface elevation 41.7 feet 

Rock floor -98 feet. Rock is typical Fordham gneiss — banded 
and very micaceous — to bottom —123 feet 
225 North side of Delancey street east of Eldridge street 
Surface elevation 40 feet 
Rock floor at el. -74 feet 

Fordham gneiss in good condition with interbedded limestone 
at bottom at el. -550 feet to bottom at el. -671 feet 
25 Delancey street between Eldridge and Allen streets 
Surface elevation 40.6 feet 

Rock floor -68.3 feet. Banded Fordham gneiss — sound rock 
— dip about 45 
233 South side of Broome street east of Allen street 
Surface elevation 42 feet 
Rock floor at el. -96 feet 

Fordham gneiss with good core recovery down to el. -200 feet 
This hole is also known as 302 under a subsequent contract 
85 Delancey street 

Surface elevation 38.7 feet 

Rock floor -82.3 feet. Banded Fordham gneiss — dip about 
6o° or less — bottom at -171 feet 
223 Grand street east of Allen street 
Surface elevation 41.2 feet 
Rock floor at el. -123 feet 
No core recovered in the first 140 feet 

Inwood limestone with dip averaging about 45° from el. -303 

feet to bottom at el. -710 feet 
Splendid core recovery 



208 Hester street east of Allen street 
Rock floor at about -145 feet 

Inwood limestone with structure at about 60 feet — 70 
Enters fairly sound rock and has continued to over 600 feet 
with dip as low as 20 , toward the bottom 
15 Delancey street near Ludlow 
Surface elevation 35.7 feet 

Rock floor -106 feet. Pegmatite and Inwood limestone, mas- 
sive and bedding obscure 
217 Southwest corner of Ludlow and Hester streets 
Surface elevation 36 feet 
Rock floor at el. -128 feet 

Inwood limestone for more than a hundred feet succeeded by 
thin strips of gneiss and limestone interpreted as inter- 
bedded Fordham 
222 Hester street west of Essex street 

Surface elevation 36.6 feet 

Rock floor at el. -130.4 feet 

Fordham gneiss with interbedded limestone showing fair core 

recovery below el. -400 feet 
Dip of rock structure about 6o° 
216 South side of Hester street between Essex and Suffolk streets 
Surface elevation 33 feet 
Rock floor at el. -167 feet 

Interbedded limestone and Fordham gneiss with a dip of ap- 
proximately 45 to el. -625 feet 
Core recovery was variable 
8 Norfolk and Grand streets 
Surface elevation 35.8 feet 

Rock floor -130 feet. Rock a close grained schistose limestone, 
Inwood, showing foliation at about 45 
231 South side of Hester street opposite Norfolk street 
Surface elevation 32 feet 
Rock floor at el. -103 feet 

Decayed gneiss and no core recovery to el. -300 feet. This 
boring was continued as no. 303 under a subsequent con- 
tract and carried to el. -525 feet with only a small recovery 
of Fordham gneiss 



218 South side of Hester street east of Norfolk street 
Surface elevation 31 feet 

Rock floor at el. -183 feet 

No core recovered in upper 300 feet 

Interbedded limestone in Fordham gneiss below el. -550 feet 
to bottom 

77 Hester street near Suffolk [see 202] 
202 Hester street west of Suffolk 
Surface elevation 30.5 feet 

Rock floor -99.5 feet. Rock all decayed to great depth 
Manhattan schist to -470 feet 
Fordham gneiss —470 feet to bottom at -577 feet 
Believed to cross fault plane 
213 Hester street 85 feet east of Suffolk street 
Surface elevation 33.3 feet 
Rock floor at el. -116. 7 feet 

The rock is Fordham gneiss of the black and white banded 
type, with dips varying from 30 to 8o°. For a very short 
distance at el. -275 feet dips of io°-i5° were recorded 
Core recovery very good 

11 Hester and Clinton streets 

Rock floor -204 feet. Badly disintegrated and no core to -279 
feet. Unusual rock, identified as a mica schist of obscure 
structure (not typical). Some calcareous portions. 
At first this was thought to belong to the Manhattan formation, 
but it is probably a schistose and rather unusual facies of the Ford- 
ham series. This hole was subsequently reoccupied and deepened 
as no. 220 under another contract with the result that an inter- 
bedded series of gneisses and limestones was shown to a total final 
depth reaching el -660 feet. Rock cores indicate dip of about 6o°. 
201 Clinton street between east Broadway and Henry street 
Surface elevation —31.3 feet 
Rock floor -133.7 f eet 

A schistose variety of Fordham gneiss with associated inter- 
bedded limestone 

219 Northwest corner of Clinton and Madison streets 
Surface elevation 26 feet 

Rock floor at el. -214 feet 

Fordham gneiss and interbedded limestone 

Good core recovery below el. -400 feet 

2 5 8 


232 Southeast corner of Clinton and Madison streets 
Surface elevation 25 feet 

This hole was reoccupied as no. 304 and penetrated the rock 
floor at el. -353 feet 

The boring has not progressed far enough to recover identifi- 
able material for rock formation 

211 East side of Clinton street south of Madison 
Surface elevation 24 

Rock floor elevation uncertain because of failure to recover 
core and the obscurity of the material washed up. Inter- 
bedded limestones and gneisses of Fordham series were 
recognized from el. -336 feet to el. -680 feet 
51 and 207 Henry street between Clinton and Montgomery 

Surface elevation 32.4 feet 

Rock floor -214.6 feet. All badly decayed to great depth mostly 
believed to belong to limestone and underlain by interbedded 
Fordham gneiss at about -345 feet 
221 Clinton street near Monroe street 
Surface elevation 22 feet 
Rock floor at el. -116 feet 

Fordham gneiss mostly very sound, with some thin interbeds 

of limestone at about el. -500 feet 
Dip of structure 45 to 8o° 
226 West side of Clinton street, north of South street 
Surface elevation 10 feet 
Rock floor at el. -37 feet 

Fordham gneiss in very sound condition showing structure at 
6o° to 90 

305 Southwest corner of Clinton and South streets 
Surface elevation 9 feet 
Rock floor at el. -50 feet 
Fordham gneiss with structure at 70 

4 Montgomery and Madison streets 
Surface elevation -32.5 feet 

Rock floor -65.5 feet. Fordham gneiss of granodiorite type 
Two borings are of special interest and significance, and because 
of the rarity of such details being recorded they are given more 
fully below. 


Each one is of great depth and indicates conditions decidedly 
different from the commonly accepted behavior for Manhattan. 

Special interpretation of hole no. 202, on Hester st. near 
Suffolk st. This is one of the very deep borings, on the pro- 
posed distributary conduit, put down to investigate the character, 
condition, and structure of the rock through which the proposed 
tunnel would pass. 

A summary of the data secured, together with an interpretations 
of conditions encountered follows : 
I Boring record (summary) 

Elevation of surface = +30.5 
a Glacial drift 

0—123 feet. Soil and various types of glacial drift 
b Residuary matter of local decay 

130-150 red micaceous mud 
c Disintegrating rock ledge too much decayed to furnish core 
150— 190 disintegration matter from pegmatite and associated* 

190-214 quartz, hornblende, chlorite, mica, disintegration sand 5 
d Decayed ledge rock capable of furnishing an occasional core 
214-224 core — several pieces of coarse feldspathic, quartzv 
mica rock 

224-237 core — several pieces of core with much green mica- 
237-255 Cuttings and disintegration sands with much gree.* 

255-277 Pegmatite cuttings 

277-305 Yellow clays and quartzose disintegration sands and 

305-314 Core-pegmatite 

314-388 Gray quartzose disintegration sands 

402-447 Coarse quartz and mica disintegration sands and finer 
quartz-mica, hornblende-chlorite cuttings that do not look 
badly decayed. The rather surprising thing is their failure 
to core 

447-463 Core — four pieces of schistose rock with white mica- 
and garnet, nearly vertical, and three fragments of pegmatite 
463-497 Cuttings only 




497-5 12 Core — a quartz biotite, feldspar schistose rock that is 
rather easily disintegrated but does not show bad 
decay. Resembles the Fordham formation more 
than the Manhattan 

512-531 Disintegration sand and cuttings containing abundant 
pearly mica 

53 '-547 Core. Many fragments of coarse quartzose and mica- 
ceous limestone, interbedded type 
e Ledge furnishing sound core 
558-559 Core from quartz vein 

573-588 Close textured quartz — feldspar — mica rock. Two 
pieces with foliation structure at about 6o° 

597-607 Typical banded Fordham gneiss with good structure. 

dip about 6o°, common black and white or gray and 
white bands in good solid condition. Thin sections 
and microscopic examination of the rock indicate 
bottom perfectly crystalline, well interlocked, fo- 
liated rock with constituents in good sound condition 

Summary of record and formation assignment 


0-123 Soil and drift 
130-150 Residuary matter of local decay 

150-500 Ledge rock considerably decayed — micaceous schist pass- 
ing into quartzose schist or gneiss mostly badly decayed, 
but occasionally giving core 

500-531 Quartzose rock resembling the Fordham rather than the 

531-547 A quartzose limestone probably interbedded with the Ford- 

558-607 Fordham gneiss, the lowermost part of which is very sound 

Discussion of meaning of this hole 
There were three rather puzzling features about the data of this 
hole at the time it was made: (1) The fact that Fordha m gneiss 
was penetrated at a point so far to the west; (2) the finding of a 
small bed of quartzose limestone in the midst of other types ; (3) the 
finding of both schistose rock closely resembling the Manhattan and 
typical Fordham gneiss in the same hole with so little space 

As to these points, the first one needs little comment. That is, it 
seems to mean that much more of this Lower East Side ground be- 


tween Madison on the east and the Bowery on the west belongs to 
the Fordham than at first supposed. This very much improves the 
outlook for safe and easy construction. 

The second one, i. e. the finding of limestone at 531 feet is prob- 
ably an interbedded limestone bed and not a part of the large In- 
wood formation. 

The third point, i. e. the finding of schists and gneisses in 
the same hole introduced more difficulty of interpretation. This dif- 
ficulty was considerably increased by the fact that the ledge is so 
badly decayed and so broken up in the drilling that no typical ma- 
terial for identification could be secured in the upper portion. There 
is no doubt as to the finding of Fordham in the lower portion. 
Later explorations support the conclusion that the whole belongs to 
the Fordham series. 

When this boring was first made, the schistose portion was. 
thought to be the Manhattan formation, and the limestone could 
then be Inwood. Subsequent exploratory work at other points has 
proven that the Fordham itself shows such schistose facies rather 
commonly where associated with the interbedded limestones. This 
is now the accepted interpretation for the whole eastern half of the 
Lower East Side belt covered in the present discussion. 

There probably is some faulting. But whether the fault 
plane dips east or west and how much the total movement is. 
has not yet been developed. This, however, is a more vital 
question than would at first appear, for if the fault dips east 
the ground to the west of it is probably all Fordham of good 
quality and will be easily explored, whereas if the fault plane dips 
to the west the whole west side for several blocks is much more 
uncertain. It is probable that the majority of the rock lying west 
of it will be of better quality than found in this hole. 

Interpretation of hole no. 207 (old no. 51) 
On Henry street midway between Clinton and Montgomery 

I )rill boring no. 207 has been put down to a depth of more than 
655 feet (approximately -633). The material is of unusual quality 
and behavior and therefore seems to require special study with a 
view to reaching a correct interpretation. The most essential points 
of the drill record are given below. 



s. Explanatory record. 
m Soil and glacial drift (surface to depth of 195 feet) 

Surface to 190 feet=sands, gravels, clays of unusual variety 
190-195 feet=reddish clay 
tb Residuary matter- — mostly decayed rock (195-247 feet) 

-212-240 feet micaceous clay — judged to be residuary because 
of the abundance of mica and the scarcity of worn quartz 
grains and rarity of foreign particles 
■c Decayed rock ledge preserving original structure 

representing interbedded limestone (247-377 feet) 
247 feet=decayed rock ledge with white blotches showing 

traces of structure 
256-330 feet=oxidized — mostly red and brown clays and sands 

from disintegration of decayed rock in place 
349-351 feet=gray micaceous clay 

251-377 feet=quartzose and micaceous disintegration sands 
and calcareous clays that effervesce in acid. Much pearly mica 
d Decayed rock leclge representing Fordham gneiss formation 
(377-489 feet) — no calcareous matter 

377-489 feet=quartz and pearly mica disintegration sand vary- 
ing from coarse to fine and mostly of very light buff color 
r Disintegration matter from a chloritized hornblendic gneiss of 
too little cohesion to witb stand the grinding action of a drill 
of so small cross section (13/16 inch). (487-532 feet) 

487-532 fcet=fine dark colored disintegration sand composed 
chiefly of quartz, chlorite and mica, the material is of same 
composition as the cores secured just below 
f Core from more substantial rock — a hornblendic gneiss sound 
enough in part to withstand the drilling process and save a 
small amount of core (532-655 feet) 

532—537 feet — 9 pieces of a green chloritic foliated rock (14 
inches-*-) structure 7o°-8o° — a close textured rock much 
oxidized and hydrated 

537-551 feet — 8 small pieces and other fragments of same 

551-566 feet — 17 pieces and several fragments sarqe rock. 
All close texture and highly chloritic 

581-596 feet — 2 small pieces (two very brown, hard pieces) 
are probably not natural — " drillite," i. e. a peculiar product 
formed by the drill when it is run too dry and partly fuses 
fragments of rock and flakes of iron from the drill into a 
-compact rocklike mass 




611-631 feet — 14 pieces same chloritic foliated rock. Two 
pieces of " drillite " 

One piece of fresh rock — a gray gneiss of rather worn texture 

646-655.5 feet — 16 pieces of — a white and black and red 
blotched rock — a garnetiferous gneiss. The rock is not a 
common type but a similar variety is sometimes seen along 
the margins of the granodiorite intrusions and belongs to the 
Fordham gneiss series. 

Rock is fairly sound and for the size of core the saving is good. 
(3 feet) 

2 Deflection test. A deflection test on this hole indicates that the 
drill lias not departed more than 5 from the vertical. 

3 Behavior of drill. It has been possible to drive the casing down 
after the drill without reducing the size and without enlarging the 
rock hole to a final depth of 625 feet. 

About half of the water fed into the machine is lost — 10 gallons 
per minute being fed and $ l / 2 gallons recovered. 

The hole filled after each pull up as much as 100 feet with mat- 
ter that either ran in from a crevice or was furnished by disinte- 
gration of the walls or was simply the settling of matters held in 
suspension during operation. These settlings or corings, as the 
case may be, were of large amount (100 feet + ) when the drill was 
cutting far below the casing and small in amount (5 feet) when 
the casing was driven down near to the bottom. This matter then 
increases as the hole is deepened again below the casing. 

Cutting and progress are rapid and easy. 

4 Examination of the rock, (a) Hornblendic gneiss. A miero- 
scopic examination of the green hornblendic gneiss shows that the 
rock is not badly crushed and that the different original grains are 
well interlocked. But the more easily affected mineral constituents 
are very generally decayed and have become especially modified on 
their surfaces where they interlock with other grains. The matter 
developed is mostly chlorite ^- a mineral that is very soft and one 
that in this case fails to furnish a very firm bond between the grains. 
A disrupting force exceeding the strength of this soft mineral there- 
fore, such as drilling with a small bit or forcing the drill, causes the 
grains one by one to roll out or break apart and furnish the sus- 
pended matter that seems to be so abundant in this hole. 

b The rock below 646 feet. This is a very unusual type of rock, 
the petrographic character of which need not be taken up here. It 
appears to be simply a contact variety, such as sometimes is devel- 



oped along the margins of the granodiorite masses where they cut 
into the banded Fordham gneiss. 

The essential feature of the rock is its fresh and sound char- 
acter. This rock is not decayed. 
5 Interpretation 

a Drift 

The glacial drift and soil cover the bed rock at this point for a 
depth of at least 195 feet. 
b Residuary soil 

Decayed residuary matter of local derivation is detected at 212 
c Bed rock 

The decayed matter still preserves the bed rock structures in a 
sample taken at 347 feet. From this point downward there 
is decayed rock ledge gradually becoming more substantial 
d Formations represented 

After bed rock is reached the first 100 feet is so altered that 
identification is not certain. At 350 feet, however, the cal- 
careous nature of some of the material is observed, and on 
this ground largely it is believed that an interbedded lime- 
stone layer is penetrated down to about 377 feet. 

From that point (377 feet) the material is very silicious and 
not at all calcareous and the core when obtained is distinctly 
gneissoid. This lower portion below (377 feet) is therefore 
judged to be typical Fordham gneiss. 

The bottom material is sound but a very rare variety for this 
e Character of contact 

Normally the interbedded limestone lies conformable to the 
structures and beds of Fordham gneiss. The structure in 
such pieces as show it indicated a dip of about 70-80 . 
Therefore the formation must stand very steep. But, so far 
as can be seen in the fragments secured, there is no direct 
evidence of a fault contact or anything abnormal. The ex- 
tremely deep alteration of the rock is the chief unusual fea- 
ture. It seems to require a better chance for water circula- 
tion than is natural in the undisturbed rock of either forma- 
tion. For this reason, I am of the opinion that there has 
been movement in this zone that weakened the rock enough 
to encourage water circulation. 

The formation dips west in normal manner at about 75 degrees. 



/ Condition of the rock 

That the upper 100 feet of ledge is very rotten can not be de- 
nied, but it is certain that this lower portion of the hole 
is not in so bad condition as the low saving of core 
would lead one to think. The grains are affected by chloritic 
alteration in such manner that they can not resist much 
disrupting force. The small diameter of drill used subjects 
the whole core to enough strain to cause the gradual pul- 
verization of the rock. This affects both the core that has 
been cut loose and the hole wall that is further subjected to 
the thrashing of the drill rods. A larger size core would 
make a very much more encouraging and fair showing. 

There may be an occasional small seam so badly decayed that 
it is encouraged to run or cave under such treatment. But 
there is absolutely no evidence that slumping or caving is 
common or even likely on any considerable scale. 

The material that partly fills up the hole when the drill i 
pulled up is believed to be in considerable part the settlings 
of suspended matter which during the agitation of drilling is 
distributed through the rising column of water. The reduc- 
tion in volume (10 gallons being fed and only 5V2 gallons 
being recovered) due to rock porosity is favorable to such 
behavior of the loosened material. 


Formations. Only three formations are represented in the 
rock floor of this section. These are the regular crystallines char- 
acteristic of all southeastern New York. 

1 Manhattan schist 

2 Inwood limestone or dolomite, and 

3 Fordham gneiss, including the Ravenswood granodiorite as a 
special intrusive member, and an unusually strong development of 
the interbedded limestones and associated schistose facies. 

These formations have their usual relation — the Manhattan 
above and youngest, the Inwood intermediate, and the Fordham 
underneath and older. These simple relations, however, are much 
complicated by dynamic disturbances of more than usual violence 
so that the series is thrown into folds so close that the individual 
beds stand almost on edge. In addition lateral thrusts of that same 
time or later have broken the strata and faulted them in several 
places. This complicates the structures still more, and, since the 



amount of displacement is in no case fully known, makes the struc- 
tures in some minor details impossible to accurately interpret at this 
stage of the work. 

Fault zones. As nearly as the material recovered can be 
classified and accredited to the above three formations it has been 
done. On this identification together with the location of points of 
greater decay the chief fault zones are drawn. The chief ones are 
judged to be thrust faults but it is possible that one is a normal 
fault. Such a combination is comparatively rare where the zones 
are so close together, but it seems to best explain the relations of 
beds as interpreted from identification of the present borings. It is 
not an unknown association though in this region. It probably in- 
dicates faulting in two different periods. This is consistent with 
the observation also that some of the fault breccia ground is not 
much decayed while others are badly affected. Probably the later 
movements have not allowed rehealing of the crevices and they are 
then the lines of chief circulation and alteration. 

It is clear, upon examination of the section as now known, 1 that 
both the eastern and western belts of limestone are too thin and 
narrow to accommodate the whole Inwood limestone. The Inwood 
normally is a formation of about 750 feet or more in thickness. It 
is therefore certain that a part of it has been cut out by squeezing 
or faulting. If by faulting then there would be expected to be in 
each case somewhat greater decay than usual along the fault zones. 
The fact therefore that such decay zones are found along one mar- 
gin of the limestone bed in each case leads to the conclusion that 
faulting is the true cause. In some cases thrust faulting would be 
required to produce the result and leave the beds standing in their 
present relations [see pi. 38]. 


The finding of limestone beds within the Fordham gneiss forma- 
tion so persistently in the Lower East Side borings is one of the 
geologically interesting and rather surprising results of recent ex- 
ploration. All of the borings in the Fordham gneiss area in this 
particular district except those near the East river have shown some 

The individual beds vary greatly in thickness, ranging from only 
a few inches to many feet. Because of the steepness of the dip of 
the beds and the obscurity of this factor in many borings it is sel- 

October 1910. 



dom possible to compute their thickness closely. It is probable that 
most of them are not over 5 to 10 feet thick, although rarely a 
thickness of 25 or 30 feet may be represented. It is certain also 
that a considerable number of separate beds are penetrated. Alt 
attempts to correlate the limestone cores from different adjacent 
holes have so far met with little success. No doubt some of those 
cut at great depth in one hole correspond to others cut higher ire 
an adjacent hole. But the differences in thickness are notable evert 
in the best cases, and it is evident that little dependence can be put 
upon uniformity of thickness as a factor in correlation. The 
foldings and crumplings. and shearing have probably affected the 
limestone members of the series more than any others. Limestones 
in comparatively thin beds are, under such conditions, especially 
liable to excessive thinning and thickening" through recrystalliza- 
tion and rock flowage. It is not at all likely that any single bed at 
present preserves much uniformity of thickness. In some places 
they are pinched out entirely while in others they may attain a 
thickness much greater than the original. It is possible also that 
some of them are repeated by folding. Whether or not this is true 
in the Lower East Side section no one can tell. On the whole there 
is no direct evidence of repetition in this way. After making al- 
lowance for all possible duplication there is still a surprisingly large 
number of limestone interbeds represented- — probably 10 — a 
larger number in succession than is known anywhere else in south- 
eastern New York [see pi. 38]. 

In petrographic character these so called limestones are all es- 
sentially very coarsely crystalline dolomitic marbles or silicated dolo- 
mites of still more complex constitution. Occasionallv a very pure 
carbonate rock is represented that corresponds in appearance very 
closely indeed to the best grades of the Inwood, but there is no doubt 
whatever of the true interbedded relation of these limestones. Their 
similarity of appearance to the Inwood in certain facies is so great 
that from the petrographic evidence alone one could not differen- 
tiate them. Their fixed relation however is unmistakable and they 
belong unquestionably to an entirely different geologic formation^ 
from the Inwood — a much older one, in fact the oldest known 
formation in southeastern New York — equivalent to the Grenville 
series of the Adirondacks and Canada. The silicated facies con- 
tains many of the common products of metamorphic processes- 
Recrystallization has produced micaceous minerals such as phlogo- 
pite and chlorite in abundance. Original and secondary quartz is 



plentiful. Serpentine, tremolite, diopside, actinolite, occasionally 
chondrodite, and rarely metalic ores are found, in many cases 
the limestone passes by transition gradually into a more and more 
silicious faciei until the rock is simply a silicious Fordham gneiss 
with quartz, mica and feldspar as the essential constituents. There 
is seldom a sharp break between the two types. Many pieces of 
apparently simple gneiss will show effervescence of a carbonate 
constituent with acid. 

The silicious beds of the gneiss series proper immediately associ- 
ated with the limestone layers are also more silicious or more mi- 
caceous than the average Fordham. They are essentially micaceous 
quartzites and mica schists and the rock generally lacks the strong 
black and white banding that characterizes the common or typical 
Fordham gneiss of other localities. It is this facies of the gneiss 
which most closely resembles certain facies of the Manhattan schist, 
and when the rock is much decayed or badly broken or is ground 
to pieces by the drill the confusion is still greater. The micaceous 
variety may readily be mistaken for Manhattan schist and the ac- 
companying limestone may equally be mistaken for Inwood. 

The occurrence of interbedded limestones of the Fordham series 
is probably more common than was formerly believed. They are 
not very often seen on the surface areas of gneiss. Possibly this 
i^ largely due to differential weathering and erosion which together 
tend to obscure those portions of outcrops where such beds maj 
occur. But the type is well known. Mr W. W. Mather in his 
Geology of the First Geological District [1843] interpreted certain 
limestones in the Highlands as interbedded in their relation to the 
gneisses there. Later workers were inclined to disregard his views 
on this point and there was a marked tendency to place all lime- 
stone occurrences in one formation. Some of the geological maps 
have been made in this way. The writer, however, raised the issue 
again in an article published in 1907 under the title " Structural and 
Stratigraphic Features of the Basal Gneisses of The Highlands," a 
N. Y. State Museum Bulletin 107. It is certain that there are inter- 
bedded limestones with the gneisses in The Highlands. More re- 
cently, the writer has recognized similar occurrences in the typical 
Fordham gneisses of The Bronx, New York city. The vicinity 
of Jerome Park reservoir is the best locality in all southeastern New 
York to see this interbedded development. The best exposures are 
at the following places. 

1 In the margin of Jerome Park reservoir at 205th street. 



2 East side of Villa avenue north of Bedford Park boulevard. 

3 East of the Concourse between 198th and 199th streets. 

4 South side of 196th street both east and west of the Concourse. 
One of these occurrences was known to the geologists of the 

United States Geological Survey [New York City, Eolio No. 83] but 
it was regarded by them as an infold of the Inwood. An examina- 
tion of all four occurrences will convince one that they are not 
infoldings. In at least two cases the structure accompanying the 
beds is actually anticlinal instead of synclinal. 

These occurrences in the vicinity of Park reservoir are 
essentially the same as those disclosed by the borings of the Lower 
East Side. In spite of its thick drift cover — 50 to 200 feet — 
there are more limestone interbeds known there than in any other 
area of similar size in tbe region. It is entirely possible that a 
thorough exploration in certain other belts might reveal an equally 
elaborate development elsewhere. 

The substantiation of interbedded limestones as a prominent 
element in certain fades of the gneiss series and their association 
with typical silicious gneiss layers with transitional relation em- 
phasizes still more the strictly sedimentary origin of at least som; 
portions of the Fordham series. Other observations lead to the 
conclusion that they are the oldest members of the series and that 
the igneous associates, of which there are many, are all younger 

One of these later intrusives is the Ravenswood granodiorite 
which cuts into the eastern margin of the Lower East Side, forms 
the floor of the present East river channel at the point of aqueduct 
crossing and continues as far as explorations have been carried into 

Structural detail of Lower East Side 

What the detailed structure of the Lower East Side is, it is im- 
possible to say at the present stage of exploratory development. Its 
general features of structure are fairly clear. The Manhattan schist, 
which is the universal floor rock of the central and western parts 
of Manhattan island, extends only a short distance east of the 
Bowery. The Inwood limestone comes to the surface of the floor 
at Christie and Forsyth streets. An anticlinal ridge of gneiss comes 
up at Eldridge and Allen streets. Then a syncline of Inwood 
limestone is pinched into the next three or four blocks and from 



this point eastward — from Norfolk street nearly to the East river — 
the Fordham gneiss with many interbeds of limestone forms the 
rock floor. 

As much of this detail as it is now possible to classify has been 
included in the accompanying drawing, plate 38, in which special 
attention has been given to the interbedded limestone occurrences. 

In view of the fact that a tunnel is finally to be constructed 
through this section which will cut the whole series of formations 
and structures at a depth probably between el. -600 and -700 feet, 
it is clear that much greater accuracy of geologic interpretation 
is soon to be attainable on many of the more obscure points. Be- 
cause of this also it is not advisable to attempt a detailed structural 
cross section at the present time. It can very well await the more 
complete data to be gathered during construction of the tunnel. 

N Y Bute Muuum Bulletin 146 

PtaM 38 

¥■ /OO 




lVo, v Jo °) 


/rom io 

7"A« lower pro/tic line it intended to 
A /Af limit 0/ rock decoy en interpreted 
at Both prominent depiemon 


one belon 

Sc/r/st- S ~ I I 
A//#esfa*r- / -AS- - — 
G/re/ss ■ Q - 1 1 
/fecenrry oyer ?5/° S/70h//7 ///i/j 
rfecore/y i/mfer " " " 
Ai? rfttcorery " '• 


£orr/tyS a/r yb/vyecfcJ />arv//e/ /b s/>-/A-e 
o/r ///re ye/mrr? Jra/e *<?fc 
/4ppn>x/mafe foirr//?? of s/rr*e /s /V ?SY 

300 Cily 0< New fork 




OCTOBER n. 1910 




Evidences of postglacial faulting and other recent movements 
have of late attracted a good deal of attention. The experience of 
San Francisco in the exceptionally disastrous earthquake and lire, 
traceable directly to earth movements of the nature of faulting 
which dislocated or injured the water conduits rendering them tise- 
lcss, is fresh in the minds of men everywhere who have public 
responsibilities of this kind. If displacements are occurring at 
the present time, or if any related movements are continuing, or if 
there is evidence of recent disturbances of this sort in this region, 
they have a decidedly important bearing upon the permanence of 
all engineering structures that cross them. 

Xo undertaking is more vitally concerned with this question than 
the Catskill aqueduct. Although the principal factors to be taken 
into account have been considered in other connections [see 
"Faults" and "Folds," pt i] a unified statement may encourage 
a more intelligent understanding of the bearing of these structures 
in southeastern New York on this specific question. 

The region included in this discussion extends from the Catskill 
mountains to New York city. It will be convenient, for the pur- 
poses of this argument, to divide the whole area into three districts 
whose boundaries are determined by decided differences in com- 
plexity of geologic history. These lines necessarily follow closely 
the boundaries of greater stratigraphic unconformities. The 
youngest groups of strata have suffered only such changes as have 
accompanied movements of later geologic periods. But before they 
were formed the underlying groups of rocks were just as pro- 
foundly affected by earlier disturbances. In this region, at least, 
three such groups of large importance exist. The oldest or lowest 
has been affected by not only everything that has influenced the 
younger strata but by disturbances of a still earlier time which verv 
much increase their complexitv. 

On this basis it is convenient to think of the three districts as 
follows : 

A Catskill district. Including that portion of the region west 
and northwest of the Shawangunk mountains and marked by the 



prevalence of Siluric and Devonic strata, i. e. all strata above the 
Hudson River slates. These strata have been affected by only one 
great mountain-making movement — that of the Appalachian fold- 
ing, and minor disturbances of still later date. 

B Hudson river district. This includes that portion of the 
region lying between the northern border of the Highlands and the 
Shawangunk mountains. It is marked by the prevalence of Cam- 
bric and Ordovicic strata, i. e. Hudson River slates, associated with 
Wappinger limestone and Poughquag quartzite as the chief bed 
rock. These strata have been affected not only by the Appalachian 
folding but also by a still earlier one — that of the Green mountains 
and the Taconic range. They were folded into mountain ranges 
and worn down in part again before the Siluric and Devonic strata 
of district A were in existence. Therefore as a structural problem 
this district (B) is approximately twice as complex as the other. 

C Highlands district. This includes all of the region com- 
monly known as the Highlands of the Hudson as well as the rest 
of the area south of the Highlands proper to New York city. Its 
rocks are the oldest — much the oldest. They had been folded into 
mountain structures and in part worn down before any of the 
others were accumulated. They have also suffered extensive 
igneous intrusion so that in places these igneous types prevail. 
And besides they have been metamorphosed far beyond the point 
of any other group. Xo other series of strata has been so pro- 
foundly affected. They form the lowest group. All things con- 
sidered this district should be structurally three times as compli- 
cated at the first one (A), and adding the igneous and metamorphic 
complexities, it is probably more near the truth to consider this 
Highland district four or five times more complex. 

All of the formations from the oldest to the Middle Devonic are 
involved. For the specific formations and their succession and rela- 
tion the reader is referred to that discussion in part I [see p. 29, 
et. seq.]. 

Structural features 

Except the most westerly part of the region, that occupied by the 
Upper Devonic strata, all formations are compressed into folds. 
Many of the smaller folds, especially those in the Catskill district, 
are still complete. The easy subdivision of strata possible in this 
district also simplifies the problem of detecting small changes of 
altitude. Rut for the most part the larger folds have been beveled 
off extensively by surface erosion so that only the truncated limbs 



are now to be seen, and the strata therefore appear as narrow belts 
that dip steeply into the ground. This is more marked in the 
Hudson river district than in the Catskill, and is still more strikingly 
true of the Highlands. 

There are evidently at least three different epochs of folding inter- 
rupting the processes of sedimentation and followed by periods of 
erosion before sedimentation was again resumed. These breaks 
constitute so called stratigraphic unconformities and occupy the 
relative positions indicated in the foregoing tabulated scheme [see 
pt 1]. 

In each epoch of folding the compressive forces accomplishing 
this work seem to have acted in a southeast-northwest direction 
causing successive series of folds with a northeast-southwest trend. 
The total amount of crustal shortening accompanying these move- 
ments is not known, but that it must be many miles is indicated by 
the fact that the strata of the older series of formations stand pre- 
vailingly on edge. All stages between small amount of movement 
to very great displacement are represented. 

Accompanying the folding in each epoch there has been a ten- 
dency to rupture and displacement of the " fault " type. There are 
multitudes of them varying from movements of too little amount to 
be regarded in a broad way to those of several hundred feet. Most 
of the larger and more persistent ones are strike faults and follow 
the main ridges or valleys, sometimes governing the location of 
escarpments or gorges. Dip faults crossing the formations also 
occur and doubtless have guided the adjustments of many tributary 
streams, and occasionally portions of the larger water courses. The 
thrust fault is most common. This is especially true of the larger 
ones and particularly those parallel to the trend of the other struc- 
tural features. 

The chief effects of these movements may be summarized as 
follows : 

1 Formations are cut out of their normal order and nonadjacent 
ones are brought in contact. 

2 Cliffs (escarpments) and sharp gulches are more common. 

3 Crush zones (belts of brecciated material) are developed. 

4 The crush zones give an additional control of stream adjust- 

All of these effects are common. Many of those faults dating 
back to the earlier epochs are obscure and not readily located. Many 
of the older weaknesses of this sort have been healed by recry^talli- 



zation so that they are now as sound as any other portion of the 
rock. A good deal depends upon the type of rock and the conditions 
under which the movement took place. In some of the more open 
ones, circulating water has seriously affected the rock and in places 
there is extensive decay even in the harder crystalline formations. 

Age of the faulting. The chief epochs of folding and faulting 
are those of the mountain-making movements — one Precambric, 
another Postordovicic, and still another Postcarbonic. All of these 
ck'te very far back in geologic history, and since the last of these, 
nothing akin to them in importance has been felt in the region. 

In Posttriassic times however there was small faulting south of 
the Highlands, that affected the areas of Triassic rocks of New 
Jersey and Connecticut. 

Whether or not there continued to be slight movement along some 
of the older lines it is now impossible to say. It is at least clear that 
all of the great movements belong to very ancient time, and that the 
last period of geologic time as we know it for this region, has been 
one of comparative stability. The chief exception is evidently con- 
nected with the continental elevations and depressions of the 
glacial epoch. 

Recent movements. The effects of glaciation make it possible 
to determine whether or not there has been further movement in 
postglacial time. Conditions are not everywhere favorable enough 
to detect minute changes, but where they do obtain, the evidence is 
capable of very definite interpretation. The essential features of 
these conditions are 

1 A bed rock ledge that has been left well smoothed by glacial 

2 Protection from postglacial destruction so that the original 
unevenness as left by the glacial smoothing can not be mistaken. 

If on such a ledge, as now exposed, there are steplike offsets or 
minute escarpments that could not have remained had they been 
present during the ice action, then there must have been displace- 
ment to this extent, since the original smoothing took place. 

A few such evidences have been found in New York and New 
England, and have been noted in geologic reports. W. \Y. Mather 
in his report on the First District of New York ( 1843) pages 156-57, 
was the first. The data as now known may be found in the last 
bulletin of Geologic Papers of the New York State Survey [see 
N. Y. Slite Mus. Bui. 107 (1907) p. 5-28]. The following para- 



graphs are intended as a brief summary and comment on the facts 
as there given : 

Localities where some postglacial displacement has been 

1 Copake, X. V., on the eastern border of the State near the 
southwest corner of Massachusetts 

2 Rensselaer. X. Y. 

3 South Troy.-X. Y. 

4 Defreestville, N. Y. (near Troy) 

5 Pumpkin Hollow, X. Y. (near Copake) 

6 Kilburn Crag. X. H. 

7 Port Kent. X. Y. (uncertain) 

8 Attleboro. Mass. 

In addition to these there is reference to similar occurrences at 
St John, X T . B. and in the province of Quebec. All of the known 
localities lie a considerable distance beyond, north and northeast, of 
the Catskill aqueduct line. 

Causes of displacement. In southern New York all of the 
cases of postglacial faulting yet discovered lie in the area of slates 
belonging to the Hudson River series. Whether the belt now occu- 
pied by this formation is therefore to be considered the most un- 
stable zone, or whether there is some tendency to slight readjust- 
ment inherent in the slates themselves causing these movements, is 
not clear. It would seem consistent with known recent geologic 
history to connect these displacements with the general elevation 
and subsidences accompanying and following the glacial occupation. 
It is. perfectly clear that the whole continental border in this region 
suffered considerable subsidence during glacial time. Also the ter- 
races and deposits along the Hudson- prove beyond question that 
during the ice retreat, at the very close of the glacial occupation, 
the land surface stood from So to 150 feet lower than now. There- 
fore an elevation of this amount has occurred in postglacial time, 
and probably, judging from the condition of the terraces themselves, 
took place soon after the glacial ice withdrew. 

The stresses and inevitable warpings accompanying these mass 
movements seem to be sufficient to account for all displacements 
known to be of this age. There is nothing in them that necessarily 
promises a renewal of mountain folding. But it appears that the 
movements liave almost all been of the thrust character and in this 
respect they differ not at all from the commoner type of the region. 



Amount of displacement. The greatest throw noted on any 
single Postglacial fault in eastern New York is given by Wood- 
worth as 6 inches, and he remarks that this is imperfectly shown. 
Usually the displacement is distributed over a zone in which several 
small faults occur instead of a single larger one. This may mean 
that the whole disturbance is essentially superficial. 

At South Troy it is stated that a total displacement of 12 inches 
is thus distributed through a number of small faults within a dis- 
tance of 30 feet. 

At Rensselaer a total of 5 inches is given. 

At Defreestville a total of 13 inches is indicated in a distance 
of 11.67 f eet - 

At Copake, at two different spots, a total of more than 7 iuches 
was measured within a space of 12 feet. Woodworth thinks that 
the total displacement for the locality may exceed 2 feet. 

At Pumpkin Hollow a total of 17 inches is estimated. 

Conclusion. If such rates prevail over larger areas beneath 
the drift, it is clear that rather profound changes would be indi- 
cated. But thus far there is no indication of such continuity. 

Likewise if it were certain that the movements are now in 
progress, it would be a matter of greater concern. But there is 
no direct evidence to prove it. 

Estimates of the length of postglacial time differ greatly. The 
shortest ones worthy of consideration range from about 5000 to 
10.000; the longest run above 100,000 years. 

Some intermediate value is probably nearer the truth — say 
25,000 years. 

Adjusting the postglacial faulting problem then to these, time 
estimates the summary of it all would be as follows : Somewhere 
within postglacial time, i. e. approximately 25,000 years, move- 
ments of strata have developed at a few places in eastern New 
York that appear as small faults with total throw in each locality 
varying from a few inches to perhaps as much as 2 feet. Whether 
the movement has been gradual and continuous or concentrated 
largely into some small portion of this time is not known. Whether 
the effects are extensive or, on the contrary, very local and super- 
ficial, is likewise unknown. But in any case there are no known 
instances of violent and large displacements, such as would be 
likely to cause great damage to sound structures, in this region in 
postglacial time. 


Appalachian mountain-folding. 63, 

66, 73- 

Aqueduct, see Catskill aqueduct. 

Arden point, 97, 104. 

Arden point line, 85. 

Artesian flows, 142. 

Ashokan dam, construction of, 13 ; 
elevation of reservoir, 17; stone 
used in construction, 38; geologi- 
cal features involved in selection of 
site for, 109-16; location map, 113; 
Olive Bridge preferable location, 
116; to be finished first, 1S3. 

Aspidocrinus scutelliformis, 42. 

Athyris spirifcroides, 38. 

Atrypa reticularis, 40, 43. 
spinosa, 40. 

Atwood, T. C, acknowledgments to, 
7; division engineer, 237. 

Beaver Kill, no. 

Becraft limestone, 42, 55, 126. 

Bensel, John A., member of Board 
of Water Supply, 13. 

Borkey, Cliarles P., consulting geolo- 
gist, 6, 19, 75 ; cited, 48. 

Binnewater sandstone, 44, 55, 126, 
133. 134. 140; porosity, 135. 

Bluestone, character and quality, 
117-23; economic features, 119; 
petrography, 119-23. 

Borings on the lower east side, tabu- 
lations and interpretations, 254-65. 

Breakneck ridge, 85, 91-95, 100, 163 ; 
quality and condition of rock, 106- 

Brink, Lawrence C, acknowledg- 
ments to, 6; division engineer, 21, 
140, 151. 

Bronx valley, geologic cross section, 

Brown, Thomas C, employed on 
Esopus division, 125; observation 
on limestone rocks. 140. 

Brush, William \\ "., acknowledg- 
ments to, 6; division engineer, 14, 
21, 215. 

Bryn Mawr, geologic section at, 205 ; 

comparison with shaft 13. 212. 
Bryn Mawr siphon. 201-8. 
Bull mountain, 163. 

Calyx drill, 26. 

Cambric quartzite, 102. 

Cambro-Ordovicic formations, 45-46, 
56, 63. 

Carson. J. P., cited, 209. 

Cat Hill gneissoid granite, 52, 57. 

Catskill aqueduct, water supply pro- 
ject, 9-16; generalities of construc- 
tion, 14-15; estimation of cost, 15; 
present plans for, 15; time for 
completion. 15 ; problems en- 
countered in the project. 17-24; 
gathering data, 21-23 > relative val- 
ues of different sources of infor- 
mation and stages of development, 
25-28 ; geologic problems. 75-276 ; 
general position of aqueduct line, 
77-80 : location map, 80. 

Catskill creek, II. 

Catskill district, general geology, 29- 
74: of simple structure, 31; post- 
glacial faulting, 271-72. 

Catskill formation, 37, 55, 63. 

Catskill Monadnock group, 73. 

( ai skill supply, area in square miles, 
11: daily supply in gallons, 11; 
estimated daily supply, 11; esti- 
mated cost, 1 1 : storage in gallons, 
11 ; part of supply available by 
79/?, 15. 

Catskill system, parts of. 11—14; con- 
struction of certain parts in ad- 
vance of the rest, 13. 

Catskill watersheds and aqueduct, 
map, 12. 

Caves, 137. 


2 7 8 


Cedar Cliff, 103. 
Cement beds, 44. 
Cenozoic time, 64. 

Chadwick, Charles N., member of 
Board of Water Supply, 13. 

Chonetes coronatus, 38. 
mucronatus, 38. 

Chop drill, 26. 

Clapp, Sidney, assistant engineer, 

Clark, cited, 44. 

Clays, in. 

Coastal plain, 73. 

Cobleskill beds, 4^, 55, 126. 

Coeymans limestone, 43, 55, 126, 133. 

Cold Spring, 85, 163. 

Conglomerates, better quality of wall 
than limestones, 140. 

Continental elevation, 67-69. 

Cortlandt series of gabbro-diorites, 
52, 53. 57- 

Coxing kill, 127, 128. 

Coxing kill section, 135-36; struc- 
tural geologic detail, 136. 

Cretaceous deposits, 36-37, 54. 

Cretaceous penpplain, 67. 

Cronomer hill, 154. 

Crosby, W. O., consulting geologist, 
6, T 9> 75 ,' cited. 36. 

Cross sections, Rondout valley, 140. 

Croton aqueduct, study of shaft 13 
and vicinity, 209-14; comparison 
of Bryn Mawr and shaft 13, 212- 
14; map showing location, 239. 

Croton lake crossing, 183-89. 

Croton river, average daily flow, 9. 

Croton water, average daily' consump- 
tion, 9. 

Crows Nest, 100, 163. 

Crystallines, south of the Highlands, 
47. 56; older, 57. 

Dalmanella testudinaria, 46. 
Dalmanites selenurus, 40. 
Dana. J. D„ cited. 48; mentioned, 46. 
Danskammer crossing, 85, 97, 103. 
Darton, mentioned, 46. 
Davis, Carlton E., acknowledgments 
to, 6; department engineer, 13. 

I Davis, J. L., tests of Kensico rocks, 

Delancey and Clinton street section, 
structural geology, 253-66. 

Delivery conduits, geological condi- 
tions affecting the location of con- 
duits, 215-29. 

Devonic strata, 37-43, 55. 

Diabase, 37, 240. 

Diamond drill, 26. 

Dikes, 106; pegmatite, 52. 

Dinnan quarry, 198. 

Division engineers, responsibility of, 

Drift, kinds of, 33-36; glacial, 32-36, 

54, 100, 202. 
Dwight, mentioned, 46. 

East river crossing, 233, 238. 
East river section, 250-53 ; structure, 

Engineers, division, responsibility of, 

Esopus creek, 11, 69, 77, 112, 128. 

Esopus division of Northern Aque- 
duct Department, 125. 

Esopus shale, 40, 55, 126, 140; thick- 
ness, 133. 

Esopus valley, geologic cross section, 

Esopus watershed, development of, 
13. 15. 

Exploration zones, special, 237-70. 
Fault zones, 266. 

Faults, 60-62. 135, 163: postglacial, 

general question of. 271-76. 
Favosites helderbergia, 43. 
Ferris quarry, 198. 
Firth Cliffe, 153. 

Flinn, Alfred D.. acknowledgments 
to, 6; department engineer, 13, 215. 
Folds, 59-60, 272. 

Fordham gneiss, 47, 52, 57, 62, 185, 
191, 192, 202, 206, 217, 218, 219, 
220, 221, 225, 226, 232, 233, 234, 
237. 238. 255, 257, 258, 260, 261, 
262, 264, 265, 266, 268. 



Formations, summary of, 54-57. 

Foundry brook, 27 ; rock condition 
at, 163-69. 

Foundry brook valley, structural de- 
tail, 165. 

Garden quarry, 198, 199. 

Geographic features, 30-31. 

Geographic history, 65-74. 

Geologic conditions affecting the 
Hudson river crossing, 97-107. 

Geologic knowledge, practical appli- 
cation to engineering plans, 19. 

Geologic problems of the aqueduct, 

Geology of region, 29-74 ! summary 
of formations, 54-57; outline of 
history, 62-65 > local summary, 265- 

Glacial drift, 32-36, 54, 100, 202. 

Glacial period. 64, 71. 

Gneisses, 176; dioritic, 198, 199; 

Grenville series, 50-52, 57. See 

also Highland gneiss. 
Grabau, A. \Y.. cited, 37, 44. 
Granites, 99, 100, 106 ; gneissoid, 

198; of the new Ferris quarry, 198. 
Grassy Sprain valley, 203. 
Gravel, no. 
Gravel hillocks, no. 
Gravel streaks, 11 1-12. 
Grenville series, 50-52, 57, 62. 

Hamilton shales, 38, 55, 78, no, 126, 

Harbor hill moraine, 35. 

Harlem river crossing, 237, 238-44 ; 

map showing plan of exploratory 

borings, 239. 
Hartnagel, cited, 44; mentioned, 154. 
Healey, John R., acknowledgments to, 

6 ; exploratory work by, 237. 
Henry street, interpretation of hole 

No. 207 on, 261-65. 
Hester street, interpretation of hole 

No. 202 on, 259-61. 
High Falls, 125, 127, 133. 

High Falls shale, 44, 55. 126, 133. '34. 

135 ; porosity, 135. 
Highlands, 30-31, 73, 81; crystal 

lines south of, 47, 56; postglacial 

faulting, 272. 
Highlands- gneiss, 50, 57, 99, 102, 154, 

163. See also Fordham gneiss. 
Highlands group, crossings, 97, 103- 

4, 105 ; more defensible as a route 

for the aqueduct line, 103. 
Hill View reservoir, 215; elevation, 


Hipparionyx proximus, 41. 

Hobbs, mentioned, 95. 

Hogan, Thomas H., assistant di- 
. vision engineer, 125. 

Hornblendic gneiss, 263. 

Hudson river, 69; water to be used 
for lire protection, 10; wash bor- 
ings, 26; depth of buried channel, 
89; submarine channel, 90-91; 
Storm King-Breakneck mountain 
profile, 91-95; origin of the present 
course, 95-96; crossing, geological 
conditions affecting, 97-107 ; out- 
line map showing possible cross- 
ings, 98; difference of structure 
in crossings, 104; postglacial fault- 
ing of district, 272. 

Hudson river canyon, 81-96; points 
of exploration, 83-88; comparative 
sections at Peggs point and Storm 
King, 92: study of profile, 94. 

Hudson River slates, 46, 56, 83, 100, 
102, 103, 126, 135, 137, 140, 149. 
153, 154, 272. 

1 Iudson schist, 201. 

Hurley, 127. 

Idlewild, 154. 
Igneous types, 52-54. 
Imperviousness and insolubility, 138- 

Inwood limestone, 47, 49-50, 56, 172, 
185, 191, 192, 201, 202, 210, 212, 
217, 218, 219, 220, 221, 226, 232, 
237, 238, 240, 242, 243. 245, 246, 
249, 254, 255, 256, 261, 262, 255, 
268, 269. 



Ithaca beds, 38, 55. 

Jameco gravels, 36. 

Jerome Park reservoir, interbedded 
development of limestones in vicin- 
ity of, 268. 

Jura-Trias formations, 37, 54. 

Kemp, James F., acknowledgments 
to, 6; consulting geologist, 6, 19, 
75; cited, 81, 232. 

Kensico dam site, geology of, 191- 

Kensico quarries, stone of, 195-200; 

additional tests, 199. 
Kensico reservoir, to be enlarged, 


Kripplebush, 127. 
Kripplebush section, 129-31. 

Laminated sand and clay, 111. 

Laminated till, HO. 

Langthorn, J. S., acknowledgments 
to, 7; division engineer, 21, 109. 

Leperditia alta, 44. 

Leptaena rhomboidalis, 40, 42. 

Leptocoelia acutiplicata, 40. 

Leptostrophia inagnifica, 41. 
perplana, 40. 

Liberty ville, 149, 150. 

Limestones, 99, 100, 176; resistance 
to solution, 139; analysis of, 139; 
of Sprout Brook valley, 171 ; in- 
terbedded, older than the Inwood, 

Liorhynchus mysia, 38. 

Little Stony point, 85, 97, 104. 

Location map, 80. 

Long Island, Cretaceous and Ter- 
tiary strata, 32; glacial deposits, 

Lower East Side zone, 238. 
Lowerre quartzite, 47, 50, 56. 

McCarthy, C. H , boring equipment 
owned and operated by, 141. 

Manhattan schist. 47. 48-49, 56, 171, 
183. 186, 191, 192, 201, 217, 218. 

219, 220, 221, 225, 226, 233, 234, 
237. 238, 240, 241, 242, 243, 244, 
245, 246, 247, 248, 249, 254, 257, 
261, 265, 268, 269. 
Manhattanville cross valley, 237, 244- 

Manlius limestone, 43-44, 55, 126, 

133, U34- 
Marcellus shales, 38, 55, 126. 
Matawan beds, 37, 54. 
Mather, W. W., cited, 268, 274. 
Merrill, cited, 48. 

Merriman, Thaddeus, acknowledg- 
ments to, 6; assistant chief engi- 
neer, 13, 109. 

Mesozoic time, 64. 

Miocene deposits, 36. 

Miocene fluffy sand, 54. 

Moodna creek, 103-4; wash borings, 
26; course of, 155-59. 

Moodna valley, ancient, 153-62; sta- 
tistics, 160. 

Morningside to Central Park sec- 
tion, 245-50. 

Mountain-forming movements, 59- 

New Ferris quarry, granite, 198. 

New Hamburg, 81. 

New Hamburg group, crossings, 97, 

102-3, 105. 
New Hamburg line, 83-85. 
New Paltz, 149. 

New Scotland shaly limestone, 42, 
55. 126, 133. 

New York-Wescchester district. 30. 

New York city, gorge at, 91 ; sec- 
tions of gorge at 32d street, 92; 
geological conditions affecting the 
location of delivery conduits, 215- 
29; areal and structural geology 
south of 59th street. 231-36; struc- 
tural geology of the lower East 
side, 253-66. 

Newark series. 37. 54. 

Ncwburgh, 154. 

Northern aqueduct to be finished 
first, 183. 

Oil-well rig, 26. 

Olive Bridge, 110; site, 112-14; pref- 
erable location for the proposed 
Ashokan dam, 116. 

Oneonta formation, 38, 55, 119. 

Onondaga limestone, 39-40, 55, 78, 
126, 127, 129. 

Oriskany beds, 40, 55, 126. 

Orthothetes woolworthanus, 42. 

Pagenstechers gorge, 154. 155, 159. 

Paleozoic time, 63. 

Palisade diabase, 37, 54. 

Pebble beds, m-12. 

Peekskill creek, 172. 

Peekskill creek valley, structure of, 

175-82; geologic cross section and 

detail of borings, 180. 
Peekskill granite, 52, 53, 57. 
Peggs point, borings, 89; gorge at, 

91, 92. 

Peggs point, crossing, 83, 97, 103. 

Pegmatite, 53, 57, 185; dikes, 52. 

Peneplain, Cretaceous, 67. 

Pennsylvania borings opposite 33d 
street, New York city, 89. 

Phyllite, 175-76, 181. 

Physiography, 30-31. 65-74. 

Piedmont belt, 73. 

Platyceras dumosum, 40. 
nodosum, 41. 

Pleistocene glaciation, 71. 

Pliocene deposits, 36, 54. 

Pompey's cave, 137-38. 

Porosity tests, 142-47. 

Porosity of Kensico rocks, 199. 

Port Ewen beds, 40, 42, 55, 126. 

Postglacial faulting, general ques- 
tion of, 271-76. 

Poughquag quartzite, 47, 56, 100, 172, 
176, 181, 272. 

Pressure tests, 27. 

Pumping experiments, 142-47. 

Quartz, 202. 

Quartzite beds, 99, 100, 176. 
Quaternary deposits, 32-36, 54. 


Raritan formation, 37, 54. 
Ravenswood granodiorite, 5 2 > 53. 57- 

217, 220, 221, 226, 233, 252, 265, 


Rhipidomella oblata, 42. 

Ridgway, Robert, acknowledgments 
to, 6; department engineer, 13. 

Rondout cement, 44. 

Rondout creek, n, 69. 

Rondout creek section, 131-35. 

Rondout siphon statistics, 141-42. 

Rondout valley section, 125-47; en- 
gineering problems, 17-19 ; geology, 
31 ; special features, 137-40; analy- 
sis of limestones, 139; cross sec- 
tions, 140. 

Ronkonkoma moraine, 35. 

Rosendale cement, 44, 45. 

Rosendale limestone, 126. 

St Nicholas Park, 246. 

Sanborn. James F., acknowledg- 
ments to, 6; division engineer, 21, 
125, 149- 

Sand, 110, in. 

Sandstones, 100. 

Saw Mill valley, 209. 

Schistose beds, 99. 

Schoharie creek, 11. 

Schoharie shale, 40, 55. 

Sedimentation structures, 58. 

Shales, better quality of wall than 
limestones, 140. 

Shaw, Charles A., member of Board 
of Water Supply, 13. 

Shawangunk conglomerate, 45, 55, 
63, 126, 127, 133, 135, 136, 149; 
thickness, 136; overthrust, 137. 

Shawangunk mountains, 31, 127. 

Sherburne beds, 38, 55, 109, 119. 

Shot drill, 26. 

Sicberell'a galeata, 43 ; figures, 43. 

pseudogaleata, 42; figures, 42. 
Siluric strata, 43-45, 55. 
Sing Sing marble, 201. 
Siphon line, total borings on, 141. 
Skunnetnunk mountain, 153, 154. 



Smith, J. Waldo, credit due, 5; chief 

engineer, 13. 
Smith, Merritt II., acknowledgments 

to, 6; deputy chief engineer, 13; 

department engineer, 14, 183. 
Smith, Wilson F., acknowledgments 

to, 7; division engineer, 21, 191. 
Snake hill, 153. 
Solubility, question of, 138. 
Southern aqueduct, general location 

map, 184; terminus, 215. 
Southern aqueduct department, 183. 
Spear, Walter E., acknowledgments 

to, 6; department engineer, 14, 215. 
Special exploration zones, 237-70. 
Spencer, J. W., cited, 90. 
Spirifer arenosus, 41 ; figures, 41. 

concinnus, 42. 

macropleura, 42 ; figures, 43. 
mucronatus, 38; figures, 39. 
murchisoni, 41. 
perlamellosus, 42. 

Sprague & Henwood, boring equip- 
ment owned and operated by, 141. 

Sprain brook, 203. 

Springtown, 149, 150. 

Sproul, A. A., acknowledgments to, 
6; division engi-uer, 21, 172, 175. 

Sprout brook, 175, 177; geology, 171- 
74; geologic cross section, 173. 

Staten Island. Cretaceous and Ter- 
tiary strata, 32. 

Stockbridge dolomite, 201, 212. 

Stony Point, 177. 

Storm King crossing, 85, 91-95, 97, 

104, 105; trial profile, 94. 
Storm King gorge, 89, 92, 156. 
Storm King granite, 52, 57, 100, 104; 

quality and condition of rock, 


Storm King mountain, fault along 

the southeast face of, 163. 
Stratigraphy, 31-57- 
Strophalosia truncata, 38. 
Stropheodonta becki, 42. 
Strophonella headleyana, 42. 
Strophostylus cxpansus, 41. 
Structural features, 58-62. 
Stviiolina fissurella, 38. 

Surface configuration, history of, 66. 
Swift, William E., acknowledgments 
to, 6; division engineer, 21, 83, 163. 

Taonurus caudagalli, 40. 
Tertiary deposits, 36-37, 54. 
Tertiary incomplete peneplanation, 


Tertiary reelevation, 70-71. 

Thirlmere aqueduct of the Man- 
chester, England, Waterworks, 138. 

Thirlmere limestone, average of five 
analyses, 139. 

Tibbit brook valley, 203. 

Till, no. 

Tompkins Cove, 177. 

Tongore site, 1 14-16; plan and geo- 
logic section, 114; detail of drift 
character, 115. 

Topographic features, 30-31. 

Tuckahoe marble, 201. 

Tuff crossing, 83. 

Uncinulus campbellanus, 42. 
Unconformities, 58-59. 

Valhalla, 191. 

Van Ingen, cited, 44. 

Veatch, cited, 35, 36. 

Wallkill river, 69, 128. 

Wallkill valley section, 149-51 ; drift 

conditions, 25. 
Wallkill-Newburgh district, 31. 
Wappinger limestone, 46, 56, 83, 100, 

102, 154, 172, 176, 181, 272. 

Wash rig, 25, 81. 
Water, increase in consumption, 9; 

reports on available sources of 

supply, 10. See also Catskill sup- 


Water Supply Board, staff, acknowl- 
edgments to, 7; members, 13; de- 
partments, 13-14. 

Wegman, cited, 209. 

West Hurley, 77, 78. 


"West Point location, 104. 
West Shokan, til. 

White, Lazarus, acknowledgments 
to, 6; division engineer, 21, 125, 

AYilbur limestone, 44. 
Winsor, Frank E., acknowledgments 
to, 6; department engineer, 14, 183. 

Woodlawn, cited, 209. 
Woodworth. mentioned, 276. 

Yonkers gneiss, igr, 106, 197, 198, 
202, 217, 218, 219, 220, 221, 225, 
226; of superior durability, 200. 

Zaphrentis prolifica, 40. 





Manhattan Schist 

Inusood Limestone 


Fvrdham Gneiss 
Yorkers Gneiss 

Rauensufood Granodiorit' 

G5 '1 \ 

Outcrops of Rock i 
Faults {cross faults) \ 


Orxgmal Lines, A-B-C^D 
New Lines rp-G-H-f J