1 7 i........ A I. .. ............ .....,w..... .......
Nk
A 'FiJA
653 3 4
A
9407 T ___________________________ 1.7
82' .p) 8.71* 78' Parto BellI
del Tra 11ndin
a, dl TroCOLO 9,Playon Grande Salud 0C
Chiriqui Chr ra NAMA oqu e
oD VID oi~enanom e'OA,,an Migue5 Palm s Aga liePERLAS SLANDS 0V ) %Sant 7
COIBA ISLAND
C)
NOTE (kal aZZone-Bound4ary line follorvs 106 ft. contour ar-oznd Gatun.-Lake.
N
y
THE
OF
CANAL
Aug,
5 4 3 2 1 0
jix.os-Sant
Sca le in Mites
0 20 40 60 80 100.
71
~1
K
)
-9.-
NrnoArrba 1% Naranjo Abajo 1
Is 9arfI Is.
MANZAN
LIMO 0f io
~ 9Irj{J~G7\A~t. ..'ewPrvsd
X7'01.Y
7-
y2 1/110 \
a /o
'HEPe
ZONE
19205
10 oKilometers"7
-qovernor The PananmChn6J
Tmbor Cay
Los Farallones Isle GsLog Niegetes cde
Jua Jaqin1.0CA
New Porto Bello RUINS ji9'
(C C 4QUARRY) Proe6
Buenavenfura Is.~
It I4'-
GordePt 1',311'
fana CblqlntoL ;z 0
Vill I'//
45, -Z tip.)9
---7 99-Ilk fill7l~ It
Mt,,,
\0 San Juan
E 1 V I g i a t S n t a
I) ~tPWdsCas&Largor
IOV I '.PDt '~ Las osc '' ":~ '~~7"~'' ~~'#ir-'' 7 4)
E ir( DrO
bra'2'' hOALR)7LI5
Pt Brujas Caame1. Pt Vique OVenado 1. Pf. Sazar *Pt Vacamonfar Tdrtola I-
5,Patterson 1.
13
A
pMelones 1. IV
0 Tabogusill
Tabco 1. QUrava I.
y
cv
e Manzanillo. CR7ISTOBAL
SAe Ncador Pf
obre de Dios
//lIh'-,,t ll
i -b ar a
liPdnain&
so' Rock
8-45- L
80*00 1
S-40'
i 14, f I I ill, _71,11. ', "." i Itz- %tlt,*
0.1 OL -
8 ei.
i 1 9.15.
I i 31, 9.00.
I Q.nn
lk
11 V
I
B-45
0
80'10'
79*45'
7TZ5'
I I
.E
1 9.-V.
Fort San Lorenz
Pifl&
0io
9*15,
9*00,
79-45-
MEMOIRS
OF THE
NATIOAL ACA DEMY OF SCIENCE-S
Volume XzVIII1
WASHINGTON
GOVERNMENT PRINTING OFFICE
1924
ADDITIONAL COPIES
OF THIS PUBLICATION MAY BE PROCURED FROM
THlE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.
AT
$1.50 PER COPY
NATIONAL ACADEMY OF SCIENCES
VOL. XVIII
REPORT OF THE COMMITTEE
OF THE
NATIONAL ACADEMY OF SCIENCES
ON
PANAMA CANAL SLIDES
III
TABLE OF CONTENTS.
Page.
Report of Committee of the National Academy of Sciences----------------------------------------------1I
APPENDIX A.-Historical sketch of the landslides of Gaillard Cut, by Whitman Cross --..---------------- 23
APPENDIX B. -The geology of the Panama Canal with special reference to the slides, by Donald F. MacDonald 45 APPENDIX C.-Chemical and physical condition of the Cucaracha, the chief sliding formation, by Warren J.
Mead and Donald F. MacDonald---------------------------------------------------- 53
APPENDIX D.-Mechanics of the Panama Canal slides, by George F. Becker-----------------------------... 69
APPENDIX E.-The movement in the slides, by Harry Fielding Reid ------------------------------------- 79
LIST OF ILLUSTRATIONS.
Frontispiece. Map of Canal Zone.
1. Panorama from southwest corner of East Culebra slide, looking north over slide and showing northern part of Gaillard Cut, December 21, 1915----------------------------..... W--------------------------- 85
2. Panorama from base of Gold Hill cliff looking northwest, showing Zion Hill, Culebra Hill, and course of canal between, December 21,1915------------------------------------------------------....... 86
3. View from Zion Hill looking south showing hilly country traversed by southern part of Gaillard Cut.----87 4. East bank of cut just south of Gold Hill, 1890, showing steep bank now broken down by Cucaracha slide.. 88 5. Cucaracha slide, near village, October 5, 1907------------------------------------------------...... 89
6. Gaillard Cut, between Gold and Contractors Hills, showing one intrusive which held back the Cucaracha slide, JJune, 911911 .. .......... ........... .......... .......... ........... .......... .......... 90
7. Cucaracha slide near Gold Hill, January 21, 1913---------------------------------------------..... 91
8. Cucaracha slide near Gold Hill, February 4, 1913---------------------------------------------....... 92
9. Cucaracha slide near Gold Hill, August 18, 1913----------------------------------------------...... 93
10. Cucaracha slide near Gold Hill, October 16, 1913----------------------------------------------...... 94
11. Cucaracha slide near Gold Hill, October 16, 1913 ----------------------------------------------..... 95
12. Cucaracha slide from west bank of canal, January 19, 1915------------------------------------------ 96
13. Cucaracha slide area near Gold Hill, December 29, 1915----------------------------------------....... 97
14. Gold Hill and adjacent area of Cucaracha slide, December, 1916----------------------------------..... 98
15. West Culebra, small slide of October 16, 1909------------------------------------------------....... 99
16. West Culebra area, shows unloading work and cut conditions, June, 1912---------------------------.....100
17. Upheaval of cut floor by movement of West Culebra slide, May 28, 1913----------------------------....101
18. Gaillard Cut, showing front of West Culebra slide, January 18, 1915-----------------------------------...102
19. West Culebra, crack near Culebra, June, 1915-----------------------------------------------------....103
20. West Culebra, crack on slope of Zion Hill, August 8, 1915-------------------------------------------.....104
21. West Culebra, view south from Culebra, August 8, 1915---------------------------------------------....105
22. West Culebra, view from Culebra toward Gold Hill, August 8, 1915------------------------------------.. 106
23. West Culebra, sunken area below Culebra, August 8, 1915 ...................---------------------------107
24. West Culebra, slide below Zion Hill, October 24, 1915----------------------------------------------....108
25. West Culebra, slide below Culebra Hill, July 3, 1916-----------------------------------------------....109
26. West Culebra, breaking up of slide on bank of canal, July 3, 1916-------------------------------------... 110
27. East Culebra slide area, unbroken face of cut, December, 1904---------------------------------------111 II
28. East Culebra slide, break opposite Culebra, June, 1910---------------------------------------------....112
29. East Culebra slide, upheaval in floor of cut, June, 1910---------------------------------------------...113
30. East Culebra slide, break, February 11, 1912------------------------------------------------------.... 114
31. East Culebra slide, break near Gold Hill, December 9, 1913-----------------------------------------....115
32. East Culebra slide, break which closed the canal, October 14-15, 1914..--------------------------------116
33. East Culebra slide, break near Gold Hill, January 16, 1915------------------------------------------....117
34. East Culebra slide, island in canal, September 21, 1915---------------------------------------------....118
-S5. East Culebra slide, broken surface of slide, October 23, 1915----------------------------------------... -119
36. East Culebra slide, obstruction of canal, October 23, 1915--------------------------------------------- 120
37. Obstruction of canal, October 23, 1915, view from surface of East Culebra slide----------------------.... -121
V
VI TABLE OF CONTENTS.
Pago.
38. Obstruction of canal by slides from both banks, November 18, 1915.................... .............. 122
39. Panorama from northern border of East Culebra slide, looking over area from Gold to Zion Hill, December
23, 1915-........ ......... ....-a -. ........ ......... ........ ........ ......... ........ ........ 123
40. Las Cascadas slide, December 15, 1910-----...---------...------------------------------------....124
41. La Pita slide (south) M ay, 1910 ............... .......... .......... ........... .......... .......... 125
42. La Pita slide (north) September 23, 1912--------------------------------------------------- ........126
43. Hagans slide, February 6, 1913---------------------------------------------------------........127
44. Hagan's slide, February 6, 1913, sunken area at rear of slide-------------------------------------.....128
45. Hagan's slide, February 7, 1913, upheaval of floor of cut---------------------------------------......129
46. Hagans slide, May 27, 1913..-----------------------------130
47. Gaillard Cut and canal from opposite Lirio slide, October 30, 1913--------------------------------......131
48. Face of Contractors Hill, December 30, 1915...................................................... 132
49. Face of Gold Hill from west bank of canal, December 29, 1915.................................... 133
50. View from Gold Hill looking north showing the widened canal in area of East Culebra slide, July 14, 1916 134 51. Small slide at Pedro Miguel ................................................... o............... 135
REPORT OF THE COMMITTEE
vu'
4
PREFACE.
Early in the year 1915 the use of the Panama Canal was dangerously threatened by serious sliding of its banks. It became necessary to give careful attention to the causes which forced movements of the earth from the excavated banks into the bed of the canal. The engineers felt that while the trouble was largely mechanical it was highly important to know what other forces might have a bearing upon it, and to what extent they entered into the problem of making the canal safe for transportation.
The President of the United States, taking advantage of the act which incorporated the National Academy of Sciences, requested the academy to consider and report upon the possibility of controlling the slides which are seriously interfering with the use of the Panama Canal." The president of the academy, at that time Dr. William H. Welch, appointed the following committee of men long experienced in geological and engineering matters: Charles R. Van Hise, president, University of Wisconsin, chairman; Brig. Gen. Henry L. Abbot, United States Army, engineer; John C. Branner, Stanford University, geologist; Whitman Cross, United States. Geological Survey, geologist; Rolla C. Carpenter, Cornell University, engineer; Arthur P. Davis, United States Reclamation Service, engineer; J. R. Freeman, Providence, R. I. consulting engineer; J. F. Hayford, director, college of engineering, Northwestern University; H. Fielding Reid, Johns Hopkins University, geologist.
The committee arrived in the Canal Zone on December 19, 1915, and spent two weeks in careful investigation. A preliminary report was published in the Proceedings of the National Academy of Sciences in April, 1916. A copy of the following more detailed report, prepared by Dr. Whitman Cross and Prof. H. Fielding Reid, was submitted to the President of the United States, Woodrow Wilson, for his consideration and approval, and shortly before the close of his term of office the report, as it is here presented, was approved for publication.
CHARLES D. WALCOTT, President.
Ix
CONTENTS.
Page.
Introduction---------------------------------------------------------------------------1.........
The problems presented to the committee. .... ... ..... .. ... .. ..... ... ... .. ..... ... .. ... .. 2
Important features of the Canal Zone---------------------------------------------------------........ 2
Topography and geography-------------------------------------------------------------........ 2
Geology---------------------------------------------------------------------------3.........
Rainfall---------------------------------------------------------------------------3.........
E arthquakes.,-.-... ........ ........ ....... ........ ........I-...... ....... -.. -.... ........ ...... 4
Development of the slides in relation to excavation-----------------------------------------------...... 5
The Culebra district and its slides------------------------------------------------------------........ 6
D om inant iim portance ........... .......... ........... ........-. ........-.. .......... ......... 6
C haracter ofof ethe lhills ............ ............ ............ ............ .. -......... ...... -..... 6
W est CC ueebra lislide .... ............ ........... ............ ............ ............ ........... 7
E ast CC ueebra lislide .... ............ ........... ............ ............ ............ ........... 8
C ucaracha sslide ............... ............... ............... .............. ............... 8
The closing of the canal in 1915----------------------------------------------------------....... 8
Causes and characteristics of the slides--------------------------------------------------------....... 9
C lassification f of ethe islides ............. ............... .............. .............. ............. 9
Primary considerations-------------------------------------------------------------........ 9
Slides oofttheirfirst aclass ........... ............ ....- ....... ............. ............ ........... 9
Slides ooftthe ecsecond asclass ......... ............ ............ ............ -............ .......... 10
G eneral pphenom ena ..... ............ ........... ........... ............ ........... ......... 10
Character of the Cucaracha formation--------------------------------------------------... I . 11
Crushing strength of the rock--------------------------------------------------------........ 11
Nature of the slide movements-------------------------------------------------------....... 12
Influence of surface water----------------------------------------------------------........ 12
Slides ooftthe hithird aclass ...... ........... ........... ........... ........... ........... ......... 13
Methods of prevention or control------------------------------------------------------------........ 13
Meth6ds tested by canal engineers-------------------------------------------------------....... 13
Suggestions from other sources - - - - - - - - - --......................... 14
Discussion of certain methods-----------------------------------------------------------........ 15
Drainage of the Cucaracha rock------------------------------------------------------....... 15
Decrease of pressure by reduction of load-----------------------------------------------....... 16
Exclusion of rain water from Culebra slides----------------------------------------------...... 16
Conclusions and recommendations-----------------------------------------------------------......... 17
A ctive sslides ..... .......... ....- .... ........ -.......... ......... .......... ......... ......... 17
Inactive slides-----------------------------------------------------------.................... 17
G eneral cconclusions ... ............ ......- .... ............ ........... ............ ......... 17
Cucaracha slide ------------------------------------------------------------------------- 18
Possible extensions of existing slides-------------------------------------------.................. 18
W est CC ueebra lislide ... ............ ........... ............ ........... ............ ......... 18
East Culebra slide ---------------------------------------------------------------------- 19
Cucaracha slide ------------------------------------------------------------------------- 19
Stability of the hills of Culebra district--------------------------------------------------------- 19
Future observation and investigation--------------------------------------------------........ 2
Importance of further study--------------------------------------------------------------- 20
Underground water and related data---------------------------------------------- .......... 20
Mechanical testing of the rock-------------------------------------------------------........ 21
Detection of movements of earth or rock-------------------------------------------.......... 21
Earthquakestustudies ..... ........ ........ ........ ........ ........ ........ ........ ......- 2
Organization of scientific work ------------------------------------------------------------ 22
General conclusions ......................................................................... 22
Figure 1. X1
REPORT OF THE COMMITTEE.
INTRODUCTION.
The committee of the National Academy of Sciences, appointed November 18, 1915, at the request of President Woodrow Wilson "to consider and report upon the possibility of controlling the slides which are seriously interfering with the use of the Panama Canal," submits herewith its final report. This report consists of a general part, which is the joint product of the committee, and several appendices, containing papers by members of the committee and others, giving the history of the canal slides, making reports of special investigations carried out at the request of the committee or discussing particular aspects of the slides.
It was desirable that President Wilson and Maj. Gen. Goethals should have the general conclusions of the coilimittee as early as possible; and, accordingly, a preliminary report was prepared, which was placed in the hands of the President on February 3, 1916. It has been published in the Proceedings of the National Academy of Sciences,.1916, volume II, pages 193-207, and in the Annual Report of the Isthmian Canal Commission for the fiscal year 1915-16, pages 587-598.
The committee as originally appointed consisted of 13 persons. For various reasons Messrs. C. D. Walcott (ex officio), G. F. Becker, and R. S. Woodward were unable to visit the canal and participate in the deliberations of the committee leading to the preliminary report. Mr. A. L. Day declined service on the committee. Those who took part in the preparation of the preliminary report were as follows: C. *R. Van uise, H. L. Abbot, J. C. Branner, Whitman Cross, R. C. Carpenter, A. P. Davis, J. R. Freeman, J. F. Hayford, and H. F. Reid.
These members sailed from New Orleans December 11 and arrived at Panama December 19. All spent two weeks in the Canal Zone, and three of them several days longer, working upon the problems submitted to them. The committee saw the canal from end to end, but directed its main attention to the slides and hills of the Culebra district, where the engineers have encountered the most serious difficulties.
The work of the committee in the field was facilitated in every way by Maj. Gen. George W. Goethals, Lieut. Col. Chester H. Harding, Lieut. Col. Jay J. Morrow, Rear Admiral H. H. Rousseau, and many other officers and engineers connected with the canal work. Gen. Goethals furnished records from his office, certain data concerning the slides and their movements, and much other information which the committee desired.
The committee has profited greatly by the geological studies of Mr. MacDonald, canal geologist from 1911 to 1913, and by conference with him in the field and in the office.
Meetings of the committee were held in Washington, D. C., April 22, 1916, and in Boston November 11, 1916, to discuss and formulate the present report. The meetings were attended by the larger number of the committee who visited the canal, and in addition Mr. R. S. Woodward attended the Washington meeting; and Mr. George F. Becker (who spent six weeks studying the slides in the Canal Zone just before the water was let into the canal in 1913) attended both meetings.
Since the preliminary report was prepared the committee has gathered much information concerning the development of the slides, secured further data from the records of the Canal Commission, and has had investigations carried out concerning certain properties of the rocks involved in the most important slides. On consideration of all available data the committee has somewhat modified the conclusions and tentative recommendations uDresented in the preliminary report.
It is a matter of regret to the committee that its study of the active Culebra slides can not now, so late in their development, be of much service so far as their "control" is concerned.
I
[MEMOIRS NATIONAL
2 PANAMA CANAL SLIDES. [VOL.
It is hoped that the conclusions of the committee regarding the slides may serve to set at rest the fears, widely entertained, as to the security of the canal and that its recommendations, though of little importance as regards the slides now active, will lead to a more thorough under-' standing of the slide phenomena and more complete future protection for this great thoroughfare of commerce.
THE PROBLEMS PRESENTED TO THE COMMITTEE.
When the committee took up the study of the Panama Canal slides, with the object of meeting the President's request for suggestions as to their control, the conditions which gave direction to their work were primarily these:
1. The canal had been closed to traffic since September 18 by the East and West Culebra slides,1 which were still active, and it was certain that they would occasion further delay before the canal could be officially reopened for commerce.
2.There was in the public mind a widespread apprehension, not felt by the canal authorities, that new slides or extensions of existing slides might cause serious interruptions to the operation of the canal for a long time to come, or even lead to its abandonment.
The problem of the moving slides was to determine whether they could be stopped or their movement lessened, to the advantage of the canal work. The problem for the future was to determine whether or not other serious slides were to be anticipated. If they seemed threatening, means of averting them must be sought for. If the canal is not in danger from future slides, a clear statement of the conditions which assure its safety must be made.
In solving these problems it was necessary for the committee to familiarize itself with the geological and' physical conditions under which the slides have occurred, to ascertain the relation between the development of the slides and, the excavation of the canal, to examine the character of the materials involved in the slides and the nature of the slide movement, to consider the experience of the canal engineers in dealing with the slides, and, to examine the existing condition of the canal banks. For clearness it is desirable to precede the conclusions and recommendations of the committee by a review of certain conditions in the Canal Zone and some discussion of the slides. Some of the data secured for the committee, the results of tests made on rocks of the Culebra district, and technical discussions of certain features, are presented in appendices, in order that they may be available for other students of the slides.
IMPORTANT FEATURES OF THE CANAL ZONE.
TOPOGRAPHY AND GEOGRAPHY.
Gaillard Cut,2 adjacent to which all slides of importance occur, traverses the higher part of the Canal Zone, from Gamboa on the Chagres River on the north (at level of Gatun Lake), through the Continental Divide, to Pedro Miguel Lock on the south, a distance of 8.75 miles. The general direction of the canal is nearly northwest and southeast. Individual stretches vary considerably from this direction, but for the sake of brevity in description, the canal will be regarded as running north and south; directions at right angles to the canal will be called east or west, and those parallel with it north or south.
The Canal Zone is irregularly hilly, with intervening areas of low relief. The views of plates 1, 2, and 3 represent the character of the country and the relations of most of the localities to be referred to. From the northern border of the Culebra slides (seen in plates 1 and 2) to Gamboa, a distance of 6 miles, or two-thirds the entire length of Gailla rd Cut, the canal passes through an undulating country with no elevations very near it attaining a height of more than about 200 feet above the bottom of the canal. The artificial banks nowhere rise more than 150 feet in this part of the cut.
I The term "slide," when unqualified, will be applied alike to landslide material which is now in motion, to that which once has been in motion butis now apparently at rest, and, in some cases, to the movement itself. Whereit is necessary to discriminate between the two conditions of the slide material, one will be called "active" and the other "inactive.",
2 "Gailiard Cut" is the name officially adopted for what was formerly called "Culebra Cut."
ACADEMY OF SCIENCESS.] XVIII.]I
REPORT OF THE COMMITTEE.
In the vicinity of Gold and Zion Hills the canal enters a more hilly country, the nature of which is shown in plate 3, a view from Zion Hill. Several hills,' the summits of which are more than 500 feet above sea, approach'close to the canal, and Gold Hill, the highest, has an elevation of 660 feet, or 620 feet above the bottom of the canal. The Continental Divide, on the canal line, was between Zion and Gold Hills at approximately 360 feet above the sea. The country traversed by the deepest part of Gaillard Cut, at the Continental Div ide, about 1 mile in length, in which all the really serious trouble has occurred, will be called the Culebra district; the hill upon which the village of Culebra stands will be called Culebra Hill.
GEOLOGY.
The geological formations of the Canal Zone are mainly stratified, sedimentary beds of shale, sandstone, conglomerate, or breccia, made up for the most part of volcanic rock debris. There are some thin limestone and carbonaceous shale layers and some of the beds contain much clay. The formations now recognized vary from a few feet to over 600 feet in thickness. The distribution of these formations is of great importance in relation to the canal.
It may be well to explain at this point that the term "rock" is used in this report in its geological sense, meaning the material of any recognized unit or formation in the earth. The idea of hardness or firmness does not enter into this definition, and, in fact some of the rocks of the Canal Zone are very soft.
In the northern: two-thirds of Gaillard Cut, north of the Culebra district, and again for nearly 2 miles at the southern, end of the cut, the rocks through which the canal passes vary in character from place to place, because of folding and faulting, and weaker rocks alternate with stronger ones in most of the vertical sections displayed in the banks. As will appear in discussing the development of the slides, the stronger rocks have to some extent supported the weaker ones in these parts of the cut, so that slides, while numerous, have been small, relatively speaking.
The Culebra district has peculiar and significant geological characteristics. Shortly north of the border of the district, seen in plates 1 and 2, the sedimentary formations assume a southerly dip (toward the point of view) and disappear beneath the canal, to reappear with a northerly dip on the southern border of the district. Thus a synclinal trough, about 1 mile wide at the canal level, crosses the ,cut directly at its deepest portion. This trough is filled by a single formation, called the Cucaracha, with a known preserved thickness of over 500 feet, and its original upper portion is no longer present.
The Cucaracha is a fine-grained, sandy, clayey formation, more nearly homogeneous throughout than most others of the region; and, as it is structurally Weak, the failure of the banks in the Culebra district has been on a larger scale than elsewhere in Gaillard Cut.
The hills adjacent to the canal in the Culebra district and to the south, seen in plate 3, are composed of intrusive basalt or a massive volcanic breccia. The breccias of Gold, Contractors, and some other hills have the appearance of being intrusive in their present position, as if forced up through softer formations by ascending basalt magma. These basalts and breccias are elements of strength in the canal banks, and their occurrence limits in various places the amount of the Cucaracha beds which can be involved in the slides. The discussion of the Culebra district and its slides will emphasize the importance of these geological relations.
The presence of a thick mantle of decayed rock and soil, such as is common in the'Tropics, has proved to be a source of much trouble in the Cucaracha slide area, where the underlying surface of soft rock is rather steeply inclined toward the cut.
RAINFALL.
The average rainfall in the Culebra district from 1884 to 1916 is reported as 87.68 inches per annum, and during the time of the excavation of the canal it was 84.75 inches per annum. This rainfall is almost wholly concentrated in eight months of the year, usually from April to
3
4 PANAMA CANAL SLIDES. [MocsNTOL
December. The average precipitation for the eight rainy months during the period of excavation of the canal by the United States was 80.01 inches.
The general effect of rain water is sufficiently summarized here by pointing out that it maintains a high level for ground water, which, occupying the large amount of pore space in certain rocks, adds materially to the load to be supported by the canal banks; it may weaken the rocks adjacent to the cut, and after slides have started their motion is accelerated by the free access of a great amount of water into the broken masses.
EARTHQUAKES.
Strong earthquakes are due to fractures of the rock and are usually connected with the formation of a new fault, or with movement on an old one. The damage done by earthquakes may be due to destruction along the fault line or to the rapid vibrations sent out. from the region of the break.
Neither the records nor the topography indicate that severe earthquakes occur in the Canal Zone. Many fault fractures were brought to light by the digging of the canal, but none seems to belong to the present geological period. Volcanic activity appears to have been absent from the Canal Zone since mid-Tertiary times, and probably the same is true of earthquakes of local origin. But, there are a number of earthquake centers not far from the zone, and shocks occurring there are often felt in the zone. The strongest recorded shock felt in the zone occurred September 7, 1882. Its origin was in northwestern South America, about 300 miles southeast of the Canal Zone-c It was an extraordinarily destructive shock in Southi America, and even on the Isthmus caused injury to the abutments and bridges of the Panama Railroad; the tracks also were said to have been bent in places. At Aspinwall (Colon) a building was damaged. No other shock has done any real damage in the Canal Zone, though a number of shocks have been felt and people have been alarmed. Los Santos Province, Chiriqui Province, and the neighborhood of Bocas del Toro seem to be the sources of these shocks. The nearest of these districts, Los Santos Province, is about 120 miles southwest of the southern end of the canal. The instruments at Balboa Heights have recorded a few shocks from nearer sources, but they have not been numerous or severe. The shock of May 27, 1914, originated about 80 miles from the southern end of the canal, probably in the direction of Los Santos Province. It made some small cracks in the Administration Building at Balboa Heights. This building has a steel frame, which is filled in with hollow tiles and plastered over. The elasticity of the steel allowed some swaying, and the brittle character of the tiles caused them to crack slightly; but the strength of the building was not impaired. The shock of October 1, 1913, caused the dislocation of a part of the side of a sluice way at the top of the Cucaracha slide.
It is well known that water-saturated ground suffers more at the time of earthquakes than solid rock, and this has aroused fears lest great slides adjacent to the canal be started by earthquakes. Neither during the times of the old and the new French companies, nor during the American regime have earthquakes exercised any observed influence on the slides; and when the slides have come to rest under a low slope of equilibrium it is hardly possible that earthquakes could do more than start a slow temporary movement. They might cause some disturbance in the railroad fills crossing swamps or to the causeway leading to Naos Island. This seems the uttermost to be apprehended. The massive reinforced structures of the canal locks make them quite secure. The comparative security of such structures is illustrated by the fact that the concrete dam of the Crystal Springs Co., about 180 feet high, situated only about a quarter of a mile from the San Andreas fault, the seat of the great California earthquake of 1906, was not in the least affected by the earthquake.
The study of earthquakes is still in its early stages, and it is impossible to say that any region will not be visited by an earthquake; but there is no reason to fear that earthquakes will do any great damage to the canal.
a Since the above was written more information has come to hand regarding this earthquake. It seems probable that it was due to move. ment on a fault under the Gulf of Darien, which extended south into Colombia and northwest under the sea passing within a hundred miles, or less, of the Isthmus. Nevertheless the conclusions arrived at in the text require no modifications.
ACDEY FSEie.] REPORT OF THE COMMITTEE. 5
DEVELOPMENT OF THE SLIDES IN RELATION TO EXCAVATION.
Bearing in mind the natural features of the Canal Zone, it is important to emphasize at this point the character of the excavation for the canal planned by the American engineers, which those in charge of the work have endeavored to execute. This original plan for the cut through the Continental Divide provided for nearly vertical walls below the water line and for'slopes of about 56' from the horizontal, or 3 vertical to 2 horizontal, above the commulnication berms following the canal a few feet above the water. The accompanying figure shows the proposed cross section for Culebra (Gaillard) Cut, taken from the Report of the Board of Consulting Engineers.3
The ability of the rocks bordering the cut to sustain such steep slopes of varying height was no doubt considered; but only the general nature of the rocks was known at the begin-. ning of excavation by the United States. Both French and American engineers made extensive, series of deep borings along the line of the canal. The detailed structure could not be ascertained until excavation made the necessary exposures. The landslides which have occurred, or which may occur, represent the modification of the canal prism by nature, i. e., the slope of its banks, in an inevitable adjustment between the differential pressures in the steep slopes left by excavation and
the weaknesses of the CLF-BRA CUT materials forming por- M -__ -__-3Z.8 t07.L0vze tions of the banks. In
the natural erosion of a
valley in sof t rocks the
slopes are modified at
the same time that the
valley is deepened. --From the history of :
the slides and the tabular ____________data concerning them FIG. 1. presented in Appendices
A and C it is evident that there has been a significant correlation between the development of the slides and progress in excavation. It appears that the slides began as soon as excavation reached a notable depth, increased in number until the bottom level of the canal had been attained in 1913; and but one small slide has begun since that time. Seventeen out of twenty-seven slides began in the years 1907-1910. Thirteen slides were in motion in the fiscal year 1908-9.
The slides have generally maintained at least sporadic activity for several years, the record showing that 18 out of 27 reached the still-existing condition of inactivity in the years 1912-1914., and but three were active on January 1, 1916.
The slides outside of the Culebra district have not followed more closely on the progress in excavation because the weaknesses resulting in slides have been of slow development in some cases. In this condition lies the basis for views expressed in a later section of this report as to the possibility of future small slides.
Within the Culebra district the slides have also developed in marked relation to excavation but under conditions not found in other parts of the cut. Owing to the thickness of the Cucaracha formation, every advance in depth necessarily increased the amount of material liable to slide, the maximum being reached only with the completion of the cut. The development of the Culebra slides has lagged behind excavation because of the time necessary for the development of a special weakness in the rocks, which is discussed in detail in dealing with these slides (p. 23).
A conception of the importance of the slide developments in Gaillard Cut may be partially expressed by some figures taken from maps, from the most recent reports, and from statements furnished by Col. Harding. In the table of Appendix C, 27 slides have received names. Onehalf of the eastern border of the cut, at the water line, has been affected by slides; the western
8Figure 1 is taken from the Report of the Board of Consulting Engineers for the Panama Canal, Washington, D. C., 1906, figure facing page 135.
[MEMOIRS NATIONAL
6PANAMA CANAL SLIDES. [VOL.
border to a less extent. The total excavation of slide material on January 1, 1917, amounted to 52,254,165 cubic yards, or more than one-fifth of the total excavation by shovels and dredges for locks, dams, and canal prism.a
THE CULEBRA DISTRICT AND ITS SLIDES.
DOMINANT IMPORTANCE.
The superior importance of the three great slides of the Culebra district over all others of Gaillard Cut, both in area and mass, is earlyy shown by the data presented in Appendix C, from which the following figures are taken.
On January 1, 1916, the acreage of the three great slides was as follows:
West Culebra slide-------------------------------- I----------------------------- 60.8
East Culebra slide------------------------------------------------------------- 70.5
Cucaracha slide........----------------------------------------------60.4
Total................................................................. 191.7
In contrast with this the total area of all other slides was 122.7 acres.
In regard to the amount of slide material removed the dominant importance of the Culebra district slides is even more marked. The figures to December 1, 1916, given in Appendix C, are:
Cubic yards.
West Culebra slide...........................................-------------I.... 11,133,428
East Culebra slide....................................................---------15,493, 828
Both Culebra slides 6---------------------------------------------------------------- 9, 961, 549
Cucaracha slide ---------------------------------------------------------- 10, 901, 999
Total------------------------------------------------------------- 47,490,804
The total excavation from all other slides to January 1, 1916, amounted to but 4,763,361 cubic yards, or about one-tenth the amount removed from the slides of the Culebra district.
The work of this committee has been concerned mainly with the great slides of the Culebra district and the igneous masses occurring in hills which limit the slides and in two places form precipitous banks rising from the canal. It is necessary, therefore, to present some details of existing conditions in this part of Gaillard Cut.
CHARACTER OF THE HILLS.
On the eastern side of the canal, Gold Hill, the highest point in the neighborhood, rises abruptly from the water to a height of 660 feet above sea level, or 620 feet above the bottom of the canal. It is composed of black, hard intrusive basalt and a mass of hard volcanic breccia, nearly surrounded by basalt. The basalt lava as it rose through the Cucaracha beds, spread out at a level 200 to 300 feet below the present summit of the hill, in what has been called "a mushroomlike form." The outer projecting parts, north and south of the summit, apparently became detached by joint fissures so that they were supported by Cucaracha beds beneath., When these supporting masses were involved in the Cucaracha and East. Culebra slides large masses of basalt and some breccia fell from the northern and southern sides of the hill. (See accounts in Appendix A.)
The present bare cliff of Gold Hill facing the canal, seen in plate 49, is an excavation face cut in breccia. A dike of basalt 20 to 30 feet wide, which formerly bounded this breccia on the canal side, has been removed down to the floor of the canal. Its extension in depth below the canal affords a buttress against lateral pressure of the Gold Hill mass.
Gold Hill separates the East Culebra and Cucaracha slides. The rugged form of the hill due to the slide breaks and canal excavation is well shown in plates 11, 18, 31, 39, and 43, together with the relations to the slides and the gorge traversed by the canal.
a Ann. Rept. Gov. of Pan. Canal, fiscal year ending June 30, 1916, pp. 308-9.
4To Jan. 1, 1916. Including the material removed from the Hodges Hill and Culebra village slides, which adjoin the active West Culebra slide on the north, and a small amount from the bank between that slide and Contractors Hill, on the south.
6 To Jan. 1, 1916. These figures include all excavation from Hagan's slide and a smaller one adjacent to the East Culebra slide, on the north.
6 For calendar year 1916; not differentiated in statement furnished.
ACDEY FSCENES]REPORT OF THE COMMITTEE. 7
On the opposite, or western, side of the canal there are three prominent hills, Contractors, Zion, and Culebra Hills, in order from south to north. Their elevations, as given on the official map of June 15, 1916, are:
Above sea Above bottom
level, of canal.
Contractors Hill------------ 415 375
Zion Hill --_------------ 550 510
Culebra Hill----------------.385 345
Contractors Hill is opposite the northern part of Cucaracha slide and its northern slope is opposite the southern cliff face of Gold Hill. Excavation for the canal cut so deeply into the eastern slope of the hill that an abrupt cliff now rises from the berm by the canal, with a maximum height of over 300 feet and extends for about 1,000 feet along the canal. The hill mass is of hard volcanic breccia (Obispo?) which is in fault contact with the Cucaracha beds about it as far as that contact has been observed. The fault rums at the base of the cliff, close to the canal, dipping westward at a high angle, varying from 650 to 80. The cliff is wholly in the breccia mass and has stood unmodified since excavation. The appearance of this fault contact as it curves away from the canal to the northern slope of the hill is represented in plate 48.
In its preliminary report the committee recommended that the crevice along this fault line, visible in plate 48, should be sealed to prevent the access of surface water; and this has been done.'
Zion IHill, the summit of which is 2,000 feet northwesterly from that of Contractors Hill, and 1,200 feet back from the canal, is represented in plate 2, in its relation to the canal, West Culebra slide, and Culebra Hill. Plate 24 reproduces the view of Zion Hill as seen from the top of Gold Hill.
Zion Hill and the somewhat larger elevation to the south (of which it is really the northern shoulder) are due to a large basalt mass, of intrusive character, as determined by Mr. MacDonald.8 On the canal side there is a marked columnar structure, shown in plate 2, which explains, in part at least, the fall of a mass of basalt when the rear border of the West Culebra slide reached the contact of the intrusive body. This structure may lead to the sloughing off of other basalt masses before this f ace of the hill assumes its permanent form.
Culebra Hill, crowned by the former Administration Building, is about 1,400 feet north of Zion Hill, the relations and appearance from the canal side being shown in plates 1, 2, and 25. This hill is due to a large mass of indurated volcanic breccia (Obispo?) similar to that of Contractors Hill. At the northern shoulder, known as Hodges Hill, is a dike of basalt, cutting the breccia or bordering it on the north. The West Culebra slide has encroached but slightly on this breccia mass, and but little more can be added to the slide from the hill side.
Much of the surface now involved in West Culebra slide was formerly covered by buildings, many of which were removed to the position now occupied on the slopes of Culebra Hill.
WEST CULEBRA SLIDE.
The active slide involves the bank of the canal for about 2,900 feet and extends westward up the slopes nearly to the summits of Zion and Culebra Hills. A large mass of basalt has fallen from the eastern face of Zion Hill, leaving a vertical cliff, whose edge is only about 50 feet from the top of the hill. (See pl. 21.) In December, 1915, a sharply defined crack extended to within 250 feet of the summit of Culebra Hill, being about 100 feet above the prominent break which was the upper limit of the slide at that time. It cuts back slightly beyond the original divide, in the narrow valleys between Culebra, Zion, and Contractors Hills, the drainage of which is toward the west. The slide extends somewhat beyond Culebra Hill to the north and ends rather indefinitely on the slope below Hodges Hill, against an older inactive slide area. On the south it is limited by a crack which runs nearly at right angles to the canal at a distance of about 800 feet from Contractors Hill, the crack being clearly represented in plate 24.
7Isth. Canal Comm., Ann. Rept 1915-16, p. 592.
8 The contact was obscured by slide material when tbe committee examined the exposures.
PANAMA CANAL SLIDES.
[MnmoIrs NATIONAL
[VOL.
EAST CULEBRA SLIDE.
The East Culebra slide is roughly rectangular, facing the canal for a distance of 2,700 feet and extending about 1,500 feet at right angles to it, reaching an altitude of 300 to 450 feet above tide. It has an area of about 70.5 acres. On the south the surface of the ground is limited by the perpendicular cliff of Gold Hill (pl. 39), a cliff formed when the slide became very active in the summer of 1915; the northern boundary is less definitely marked. On the east the slide extends slightly beyond the subordinate divide, so that the drainage of the land immediately east of the slide is away from it. The country north and east of the slide is rather flat as may be seen from plates 1 and 2.
Movement in this slide area probably began in 1906. As excavation progressed slides increased in number and size until the' final unification of the whole in one great slide in 1915. The views of plates 1, 2, and 28 to 39 show the appearance and relations of East Culebra slide in several stages of its development.
CUCARACHA SLIDE.
The Cucaracha slide, including its newer and older portions, faces the canal for a distance of 2,800 feet, and extends eastward, diminishing in width, 1,900 feet from the canal's axis, and reaches the subordinate divide beyond the crest of Gold Hill; it has an area of 60.4 acres, and drains an area of about 80 acres. Its highest part rises more than 500 feet above the bottom of the canal. Except on the side of the canal the Cucaracha slide is nearly surrounded by hills of intrusive rock, seen in plates 12 and 13.
Movements have taken place in parts of this slide area since tue time of the old French company, and have continued with intervals of rest, as the prism of the canal was deepened, but the whole area of the slide has never been active at one time. The early movements were of superficial material. (See pl. 5.) As the cut was deepened rock of the Cucaracha formation became involved, but was held back for a time by a dike of intrusive rock near the canal. -In January, 1913, this dike broke and a great slide poured into the prism of the canal. (See PS. 8-11.) Not only was the sliding material which entered the prism removed, but a large additional amount was removed from the bank, leaving a broad, fiat area which would hold a large part of the slide if it should again advance at this place. (See ps. 12 and 13.) There has been no renewed activity of practical importance in the slide proper since 1913. In August, 1916, an accumulation of loose material on the bank close to Gold Hill slipped into the canal, blocking it for a few days. (See Appendix A, p. 42.) THE CLOSING OF THE CANAL IN 1915.
The canal was opened to commerce in August, 1914. The first interruption to its navigation came in October of the same year when a section of the east bank measuring, 2,000 feet along the canal and 1,000 feet back from it gave way, settled as much as 20 feet in some places, and squeezed enough material into the canal to close it to navigation for about one week. A second interruption occurred, from October 31 to November 4, but the subsequent movement of the slide was slow and intermittent, so that dredges were able to keep the channel clear until August, 1915. At this time a sudden advance of a part of the West Culebra slide took place, closing the canal for a few days. Although some ships were able to pass through the canal after this slide, navigation became uncertain and the combined movements from the east and west banks soon exceeded the capacity of the dredges for removal.
On September 18, 1915, the canal was officially declared closed. From that date until December 18 no ships of considerable size were able to traverse the canal. Even after the slide barrier had been cut through, the depth of the channel was uncertain from day to day, and the canal authorities did not consider the conditions satisfactory for the resumption of traffic until April 15, 1916, when the canal was officially reopened.
Gen. Goethals has described and illustrated the conditions under which the canal was closed in a Supplement to the Canal Record for January 5, 1916, and again in his annual report for 1915-16.1 In the "History of the slides," Appendix A, will be found further details of the slides which closed the canal.
Ann. Rept. sth. Canalcor., fiscal year ending June 30, 1916, pp. 27-42. A general review of the slides.
8
ACADEMY OF SCIENCES.)
REPORT OF THE COMMITTEE.
CAUSES AND CHARACTERISTICS OF THE SLIDES.
CLASSIFICATION OF THE SLIDES.
On reviewing the origin and character of the slides it is found that they are subject to an appropriate classification, which it is desirable to follow more or less closely in the succeeding discussion. This classification is as follows:
Class 1. Movements of rock masses on a preexisting plane or zone of weakness, inclined toward the canal. Class IL. Movements along newly-formed fractures or by flowage or the two combined.
These slides, which are of the greatest importance, are -now called breaks by the canal engineers. Class 111. Movements of surface material:
(a) Soil with associated rock debris and vegetation.
(b) Dump material, talus, etc.
Primary considerations.-The excavation of the canal cut was the primary factor in the origin of all the slides, through the new play given to the force of gravity in its banks. However, the failure of the banks under this force was due to inherent weaknesses of materials, leading to the slides under various local conditions. In this development of the slide movements surface water has been an important agent, of special efficiency on account of the heavy rainfall of the region. Water acts in different ways in the three classes which have been distinguished.
For the sake of brevity the various weaknesses of the rocks, the agencies which have developed them, and the characteristics of the resulting slides, will be considered together in discussing the several kinds of slides.
SLIDES OF THE FIRST CLASS.
The planes or zones whose presence in the rocks produce an inherent weakness which may be developed, particularly through the influence of percolating water, to the point of causing the slides of Class I, are of several kinds. They may be primary structural features or fractures of secondary origin. Among the former may be specified: (a) The upper surface of a rock which is relatively impervious to water, on which rests a pervious rock or group of rocks; (b) a porous, weak stratum beneath a dense or less porous one. The fractures may be fault fissures or common joint planes. Where any of these planes or zones are inclined toward the canal and intersect its banks and water has access to them. the conditions are favorable for slides.
The influence of water circulating along fault planes or zones. is to soften the clay or fine attrition material commonly found bordering these fractures and to act as a lubricant for any movement which occurs. It softens, weakens, and may even render plastic the material of a layer or contact zone supporting a heavy load.
The detailed studies of Mr. MacDonald have shown that the conditions above briefly outlined occur in many parts of Gaillard Cut north of the Culebra district. Several slides observed by him clearly belong to Class I, and others, which took place before or after his connected service as canal geologist, were apparently of the same character.
The lower La Pita slide, on the east bank 2" miles north of Gold Hill, described in Appendix A and illustrated by plate 41, shows how conditions referred to above may combine to produce a slide. In this case water from the East Obispo diversion ditch seeped down along two steep fault planes to an inclined zone, and there softened and weakened to the point of failure a porous stratum which supported an overlying breccia mass. When the porous rock finally gave way the breccia slid into the cut.
The committee had no occasion to examine any particular slide of this class, for all are now inactive; but the committee will call attention, where making its recommendations, to the desirability of continued watchfulness in the region where they have occurred.
It was thought by Dr. Hayes and the central division engineers that some of the earlier slides in the West Culebra area were due to the presence in the Cucaracha formation of thin,
9
4
10 PANAMA CANAL SLIDES. [MVVOSNTOL
slippery carbonaceous layers which dipped gently toward the canal prism and disappeared below its floor. The rock above this horizon is said to have slipped without disturbance of the material below the plane of movement. Owing to later movements involving deeper portions of the Cucaracha formation, it is not now possible to confirm this idea.
SLIDES OF THE SECOND CLASS.
General phenomena.-Slides of the second class have distinguishing features of great importance which should be understood by engineers or others who would suggest practical plans for their control. The stages in their development are striking and peculiar to them. The first indication of a slide of this class is the opening of one or more deep and nearly vertical cracks in the ground at some distance from the canal. It may be weeks, months, or even years after the first appearance of the cracks before important slide movements take place. Then the ground immediately in front of the cracks sinks along a slightly curved and, at this place, almost vertical surface; contemporaneous, or nearly so, with this sinking there is marked lateral movement into the cut near or at its bottom followed by breaking down of the banks of the cut. In some cases the floor of the excavation has been forced up vertically making a mound, or in some cases, after water was admitted, causing an island in the canal and blocking the channel. Evidently the cracks, the subsidence, the lateral movement, and the upheaval are all related phenomena. Between the rear crack and the cut the ground gradually breaks up, portions near the crack exhibiting marked subsidence and those near the cut a greater lateral movement toward it. Gradually the whole mass moves toward the cut, slowly, with varying velocity, and with periods of rest.
Many features of these slides are graphically shown in the plates accompanying this report, and descriptions are given in the historical sketch, Appendix A.
The inner or basal limit of the slide is believed to be a strongly curved surface near the upper end of the slide, and in some cases also near the lower end. Intermediate portions may possibly be slightly curved.
The first recorded slide of this class occurred in 1907. The break was near the cut, when the bank was but 100 feet high and the mass of the slide was not large enough apparently to arouse especial apprehension. The earlier slides of this class involved hardly as much as 100,000 cubic yards; but as the canal cut became deeper, the breaks occurred at a greater distance from the central line, and the area and volume of the slides increased enormously. Originally distinct slides were finally connected, until in 1911-12 the East Culebra slide covered an area of 50.7 acres and the West Culebra slide covered an area of 63 acres (including Hodges Hill and Culebra village slides); at the end of 1915, the three great slides had a combined area of nearly 192 acres.
The cracks presaging large slides have appeared successively, farther and farther from the cut, as depth of excavation increased. In the West Culebra slide area the first observed vertical crack formed in 1906 and was only about 100 feet from the edge of the canal cut. By 1909 cracks had formed from 600 to 900 feet back from the center line of the canal prism and extended nearly 2,'000 feet parallel with the canal. By 1910 large cracks had extended so far up Culebra Hill as to require the removal of a number of buildings. In 1914 cracks extended from the upper slopes of Culebra Hill to the front of Zion Hill. Plates 19 and 20 illustrate the nature of these cracks.
Sudden subsidences adjoining cracks have been notable in many places. The formation of cracks in West Culebra slide was often marked by notable subsidence, and at the time of the great movement of 1913, large masses of Culebra and Zion Hills fell. In 1911, a section of the East Culebra slide, bounded by a crack nearly 1,100 feet long, sank vertically about 30 feet. In 1913, a subsidence of 60 feet occurred in the northern part of the slide. Many instances of such movement are given in the historical sketches. An illustration of such a zone of subsidence is' given in plate 44.
The contemporaneous horizontal movements in the banks of the cut were equally important; a horizontal movement of 80 feet accompanied the subsidence of 60 feet in Hagan's slide in 1913.
AcAiDnmy oF SCIENCEs.]REOT FTH CM ITE.1 XVIII.] RPR FTECMITE
Upheavals in the bed of the cut were impressive and the connection with subsidence was evident. The first upheaval was during the slide of October, 1907. This upheaval continued throughout the entire period of dry excavation, until it finally ceased in 1913. In length these upheavings varied from 50 to 150 feet, with widths varying from 15 to 100 feet, all practically coincident with a settlement in the banks back from the face of the prism.'0 By 1910 a stretch of the bottom 2,000 feet long north of Gold Hill was affected in this way, but not all at once. In 1913 there was an upheaval 30 feet high, and in October, 1915, the bed of the canal rose 45 feet and an island appeared where there had been deep water. The maximum recorded elevation of 65 feet, attained in one hour, took place about this time.1' The mounds- of upheaval are illustrated in plates 17, 29, and 34.
Character of the Czcaracha formation.-The explanation of these slide movements must rest on an understanding of the character of the Cucaracha formation, the principal material involved in them. In Appendix B, on the geology, Mr. MacDonald presents the details of present knowledge of the Cucaracha rock, and hin Appendix C the results of various tests of its properties are given. The more significant features bearing on the slide problem will be summarized here.
The Cucaracha rock was originally a stratified sandstone the grains of which were chiefly particles of volcanic rocks or of their constituent minerals, with a considerable admixture of quartz and other minerals of a different origin. These grains were very small, averaging 0.3 to 0.7 millimeters in some specimens, and rarely exceeding a few millimeters. While there is some variation in character of materials and in size of grain, from layer to layer, the formation has no specially strong beds and is of pronounced weakness throughout.
This fine-grained sandstone has been very greatly decomposed in some past epoch, all of its constituents but the quartz being replaced by or altered to secondary minerals in much of the mass. The nature of the original volcanic rock grains can not now be accurately determined. They were rich in iron but were probably not as basic as basalt. The resulting minerals are chiefly kaolin, mica, chlorite, and iron oxides. The chemical analyses of green and red phases are given in Appendix C with calculations of a corresponding mineral composition, which agree with results of microscopic study, so far -as the extremely minute and obscure particles can be recognized. The red, oxidized rock of the upper zone and some lower beds, and the prevailing gray or greenish material are of nearly the same character except for the red iron oxide which causes the color of the former.
The Cucaracha material, unlike many altered rocks, has not sufficient cementing substance to give much firmness and strength. On the contrary, it has been found to contain a large amount of pore space and these pores are filled by water below the ground-water table, which can not sink far below the surf ace, even in the dry season. From tests made by Mead and MacDonald on 21 specimens it appears that the loss of water on drying at 1000 C. is equivalent to nearly 28 per cent of the volume of the rock. It has been found that the pores are for the most part of capillary size and the water itself -may to some extent play the part of matrix or cement in the rock. A further discussion of the pore space and water content of this rock is given in treating the problem of drainage, page 61.
Crushing strength of the rock.-Samples of the typical phases of the Cucaracha formation were collected in their natural saturated condition from below the ground-water level, sealed with water in water-tight cans, and shipped to the Bureau of Standards at Washington for crushing strength tests. The blocks were sawed into test specimens and tested in their natural saturated condition. The results of these tests are discussed by Mead and MacDonald in Appendix C. The specimens tested represent the harder and firmer phases of the formation, because only such phases would yield material which couldbe sawed into blocks and tested. Of seven specimens tested, the strongest f ailed under pressure of 865 pounds per square inch. Two failed at slightly over 800 pounds per square inch, three between 400 and 500 pounds per square inch, and one at 280 pounds to the square inch. Some of the specimens were not of
10 Letter of Gen. Goethals to Chairman Van Hise, dated Jan. 5, 1916.
11 Gen. Gothals in Isth. Canal cor. Ann. Itept., 1915-16 p. 32.
12 PANAMA CANAL SLIDES. [ ViooILNA .1
sufficient strength to permit the preparation of test specimens. Consequently it may be safely said that some phases of the Cucaracha formation have a compressive strength of less than 200 pounds per square inch. The strength of the Cucaracha formation as a whole is determined by the weakness of the weaker beds rather than by the strength of the stronger beds.
During the compressive tests on the rock, water was observed to be squeezed out of pores and minute cracks in the specimens at pressures as low as 100 pounds to the square inch, although these specimens did not actually f ail during the test until pressures of 400 to 800 pounds to the square inch were reached. The stress deformation curves representing these tests indicate that f ailure actually began in most cases at pressures varying from less than 100 to 200 pounds per square inch.
Nature of the slide movements.-What are the conclusions which should be drawn in regard to the nature of these movements which are-marked by breaks, subsidence, lateral movements, and upheaval? The nearly vertical subsidence of a block of ground along the upper border of a slide area can be explained in but two ways. The rock beneath the sunken area has been either condensed in volume to an amount equal to the subsidence, or displaced laterally to the same extent. The first alternative may be dismissed as practically impossible; the second must be accepted as correct, because the canal cut has provided the space into which the lateral movement of material can take place, and such movement has occurred practically contemporaneously with the subsidence.
If subsidence at the upper, and lateral (horizontal) movement at the lower, limit of the slides are parts of one larger movement, the nature of the necessary rock deformation between these extremes, which is not open to observation, is a matter of both theoretical and practical interest. The yielding of the rock in depth, under existing conditions, may be interpreted as a plastic flow-that is, a deformation without rupture, or as rupture without plastic flow or a combination of the two. The subject is discussed more technically in Appendixes D and E.
In accordance with any one of the three interpretations and in harmony with observation the slide movement is nearly vertical near the upper limit, curves strongly toward the canal as depth increases, and approaches the wall of the cut almost horizontally. Apparently the stress on the rocks, which is exerted for a short distance below the bottom of the cut, met by the resistance of the opposite bank, has been relieved, in some cases, by a buckling up of the floor-the so-called "upheaval" or "bulging." Where this occurs the outer limit of the slide at this end is strongly curved upward.
Whether or not the deeper movement is one of flowage, the upper parts of the moving mass, which are not under a crushing or deforming pressure, are broken and fractured throughout. The exposed surfaces of the slides are very rough on account of the breaking up of the rock, as illustrated, particularly in plates 25, 26, and 35.
Influence 'o surface water.-So long as the Cucaracha formation of the canal banks sustains its load-that is, until the slide movement has begun-the rainfall of the region can have no direct effect in promoting the development of slides, except as it helps to maintain the high level of ground water. It is a matter of record that the early movements of the slides of the Culebra type have occurred at all times of the year and have not borne any special time relation to the rainy season.
On the other hand, after a slide has commenced, the broken surface, and more especially the deep cracks at the head of the slide, may take in much water, and this may be kneaded into the ground by the movement, supersaturating the mass locally, and developing places of special weakness, which may accelerate the movement. It is probable that most persons who have thought on this question have greatly overrated the direct effect of surface water in promotmng the movement of the Culebra slides and have underrated the importance of ground water or have ignored it; but rain water probably increases materially the mobility of the slides after movement has begun. In view of. their enormous masses and the difficulty hitherto experienced in making removal by dredges keep pace with advance of the slides at the canal, the problem of excluding rain water from the slides is, or has been, at least, of particular interest and is discussed later on as a means of control.
ACAIM FSac~.] REPORT OF THE COMMITTEE. 13
SLIDES OF THE THIRD CLASS.
The slides of Class III require but brief mention because their character is well known and there is no likelihood of serious trouble in future from such movements. The early slides of the Cucaracha area were mud flows of this class, which have caused much trouble at times, as is described in the "History of the slides." The basin south of Gold Hill, where the slope toward the canal cut was unusually steep, has been the principal scene of these flows and a large part of the surface material of that basin, liable to such movement, has already slid into the cut and been removed.
Mud slides of the Cucaracha type have characterized the rainy season after the water had rendered some mass fluid and where excavation had removed the hindrance to such flow. The behavior of the rocks on being wet by rain after drying during the dry season, greatly favors mud flows. (See Appendix A, p. 26.) The cementing material of these rocks appears to lose part of its water of hydration on drying at ordinary temperatures. The rock shrinks and a multitude of shrinkage cracks develop. The individual blocks of the dried rock bounded by these cracks are firm and hard. However, the effect of water on the dry rock is to cause most complete disintegration. The rock breaks down to such an extent that each individual grain is actually separated from its neighbor and the result is a fluidlike mud totally devoid of coherence, which flows readily.
Movements of dump material, incautiously deposited on slopes inclined toward the canal, have been reported from the Cucaracha, and East and West Culebra areas. In some cases this spoil of excavation has slipped with the superficial mantle on which it was deposited and in other cases it is said to have moved on a surface of soft rock lubricated by percolating rain water.
It is considered unnecessary to discuss in this report the slumping of loose material, especially when watersoaked, in railroad embankments, dams, breakwaters, or dumps of quarries. Such movements have occasioned much trouble during the construction of the canal and its associated works, but they are common phenomena attending large engineering projects and their nature is well understood.
METHODS OF PREVENTION OR CONTROL. METHODS TESTED BY THE CANAL ENGINEERS.
Ever since the slides began to cause trouble the engineers in charge have made efforts to control or prevent the development of the slides. Most of the effort has been to stop or retard the progress of active slides. Little attempt has been made to prevent slides by any method applied before initial movement had taken place. In the case of slides due to some structural weakness (Class I) they occurred before the geological structure and the effect of percolating water had been studied. In slides of the Culebra type the formation of long cracks preceded the lightening of the load on the banks.
The French engineers had to do only with surface mud flows such as those which occurred at the Cucaracha slide at an early date. They attempted to drain affected areas by surface and subsurface drains of various kinds and by tunnels. These efforts were in the main unsuccessful. The conditions to which those methods were applicable are not duplicated in the slide areas of the present time.
In his annual report for 1915-16 Gen. Goethals has reviewed the various methods for the control of slides, tested under his direction,'12 and similar descriptions were given personally to the committee. The absolute inefficiency of most of these methods makes a brief mention all that is necessary.
In four different instances piles were driven into slides with the hope of reaching firmer ground below and thus holding back the sliding material, but the piles moved along with it. In some places heavy stones were placed on sliding clayey material with hopes that they would
12 Istli. Canal Cor. Ann. Rept. fiscal year ending June 30, 1916, pp. 36-40.
14 PANAMA CANAL SLIDES. [MEO VSN OL
sink to the bottom and stop the movement; this also was unsuccessful; and many of the stones were carried into the prism and later removed by the shovels.
Attempts have been made, indirectly, to strengthen the Cucaracha formation by preventing the rain water from saturating the ground. Planting 'slopes with vegetation has not seemed to affect the -slides. This process is still used to prevent surface wash, with some measure of success. The application of concrete to the walls of the cut has been tried in some places, rather to prevent surface disintegration than slide movement. It has not been successful.
The American engineers have -made no systematic efforts to keep water away from the slides or threatening ground. Drainage ditches have been at times opened down the slopes into the canal, on both sides as far south as Gold H1il, and also on the Cucaracha slide when it became inactive; but it was formerly considered impracticable to maintain open drains on the active slides. In the preliminary report of this committee the establishment of a system of drains on the great slides of the Culebra district was recommended. Considerable new work in this direction appears to have been done according to the recent report of Mr. W. G. Comber 13 who states that:
Ditches were excavated at Culebra slide east, Culebra slide west, and at Cucaracha by the hydraulic graders, to provide a quick run-off for the storm surface waters, and prevent ponds from forming in the low areas and depressions. Over 7,500 linear feet of ditch were excavated and maintained.
The east and west diversions are large ditches roughly parallel with the canal dug before the Culebra slides began for the purpose of diverting streams from the cut during the period of dry excavation.
The prevention and retardation of slides by reduction of pressure on steep banks through removal of material from the upper slopes has been tried in several places. When the Cucaracha slide started afresh, in 1913, 1,000,000 cubic yards of material were removed by sluicing from its head away from the canal slope. In recent years the hydraulic grader has been used with good effect in reducing the slope of threatening areas adjacent to the Culebra slides. It was also used in January, 1916, to reduce the slope of the Buena Vista slide, on the west bank of the canal in the northern part of the cut. This small slide, which began in 1908, renewed its movement in December, 1915. The grader was put at work in January, 1916, to reduce the slope, and movement soon ceased.
Reduction of pressure in the Culebra slide areas by steam-shovel work was recommended by Dr. Hayes in November, 1910, in his report to President Taft.'14 The work of lowering the banks was undertaken in January, 19 11, and continued until December, 1913. The appearance of the West Culebra slope in June, 1912, and the manner of prosecuting this work are very clearly exhibited in plate 16. More than 6,500,000 cubic yards were removed from the east bank and nearly 8,800,000 cubic yards from the west bank, leaving slopes between 1 on 1.5 and 1 on 6.5 on the east side, and between 1 on 2.46 and 1 on 4.35 on the west. This reduction of the pressure was thought to have stopped the slides, but renewed movement, following a deepening of the cut, led to further shovel work in 1914. Work of this kind was abandoned after water was admitted to the cut.
SUGGESTIONS FROM OTHER SOURCES.
A general interest in the success of the canal has led many persons, engineers and laymen, to offer suggestions for the control of the slides to the canal authorities and to this committee. Most of the suggestions have reference to the Culebra slides and it is certain that most of them would not have been offered had their authors possessed any clear conception of the magnitude of the slides, the character of the rock involved, the nature of the movement, or of the deep deformation to which it is due.
The suggestions for resisting the flow or strengthening the rock directlyby the methods indicated below f ail to take into consideration the volume of the slide or the depth of the movement. No solid foundation could be found for revetments; large and deep solid columns of concrete
13 Isth. Canal Com. Ann. Rept. 1915-16, p. 305.
14Canal Record, Vol. IVI Dec. 7? 1910, P. 115.
ACADEMY op SCIENCES.]RPR FTE O MTE.1 XVIII.]REOTO TH COMT E.1
could not be constructed in moving ground; or, if this were possible, they would be carried down in the slide. There are not sufficient open fissures into which cement grouting could be forced, except near the surface. Freezing of the ground could not be carried to a sufficient depth, even if such a precarious remedy could be taken seriously.
There is a widespread belief that if water could be prevented from entering the ground and if the water already there could be abstracted the slides would cease.
In order to keep the water out of the ground, it has been proposed to smooth the slopes rising from the Canal and then cover them with some material impervious to water, such as oil, tar, bitumen, or asphaltum. Cement also has been suggested, either laid as a continuous sheet, or projected over the surface and mixed with the surface material by means of special devices.
To remove the water already in the ground, it has been proposed to drive tunnels to act as drains, and to sink a system of wells, pumping the water from them continually.
of The practicability of removing or excluding water from the slides or undisturbed masses ofthe Cucaracha rock will be discussed in the next section of the report.
DISCUSSION OF CERTAIN METHODS.
Some of the methods for preventing or controlling the slides, tested by the canal engineers, suggested by other persons, or discussed in the preliminary report of this committee, are worthy of further consideration. These methods will now be reviewed, in the light of some new data regarding the Cucaracha formation, to determine whether they can be expected to accomplish certain results and whether, under present conditions, it is desirable to secure those results.
Drainage of the Cucaracha rocl.-It is a fairly general belief that if a considerable part of the ground water could be removed from the Cucaracha rock of the unbroken banks its crushing strength would be so materially increased that the liability to slide movement would be decreased. In the preliminary report of this committee it was suggested that experiments be made in tile and tunnel drainage and that tests of porosity, water contents, and mechanical strength of the rock be carried out.
At the instance of the conuittee, experimental studies of the pore space and water content of the Cucaracha formation have been made by Mead and MacDonald. (Appendix C.) These studies show that the rock contains a large percentage of pore water. In 21 specimens taken from below the level of ground water, and therefore completely saturated, the average water driven off at 100' C. was 12.2 per cent by weight, equivalent to 27.8 per cent by volume. Sixteen specimens of ISimilar character taken well above the level of ground w ater, where the rock is jointed and fractured throughout, and therefore under conditions of perfect drainage, were found to contain 10.6 per cent of water by weight. The decrease in water contents under conditions of perfect drainage was, therefore, only 13 per cent, or less than one-seventh of the water contained in the saturated rock. Not all of this loss can be attributed to drainage, for a small part, at least, was due to evaporation. Assuming that these figures apply to the whole mass of the Cucaracha formation it is probable that less than 13 per cent of the ground water can be removed by drainage.
The above facts make it clear that the material of -the Cucaracha formation is so fine grained, that while the total volume of pore space is very large, each individual opening is so small that the water is firmly held by capillary attraction.
It is manifest that no method of drainage will remove the water from more than a very small portion of the pore spaces, even near the surface of the slide. The only water which can be effectively removed is that in the fracture openings. But the pressure in the deeper parts of the great slides will prevent the formation there of fracture openings of sufficient size to serve for the transmission of more than a small amount of water.
Whatever result might be obtained by drainage would be in part counterbalanced by water drawn from without the area of drainage by capillary attraction. The deformation of the rock extends below the level of the water in the canal in many cases-that is, below the level of possible drainage.
Ak
16 PANAMA CANAL SLIDES. [MEORSNTONL
Decrease of pressure by reduction of load.-This method of preventing the formation of new slides or the extension of existing ones has now small importance as compared with what it had in early stages of the canal excavation, but it is still worthy of discussion. There have been differences of opinion as to the relative cost and advantages of removal of slide material at the cut and of unloading above and thus preventing a slide. The solution of this problem naturally requires some knowledge of the occurrence, character, and strength of the material involved and the nature and probable extent of the anticipated slide movement. One must be able to estimate approximately the cost of prevention as compared with that of removal of slide material.
With slides of the Culebra type several factors favor prevention by unloading or reduction of slope. The amount of rock which is involved in a slide is much greater than the amount, the timely removal of which would have prevented the slide. The maximum load to be removed for safety is not much greater than the excess of existing load above that which the rock can sustain. A thorough study of the problem would give the basis for at least an approximate solution. Some suggestions are given in Mr. Becker's paper, Appendix D.,.
The damage, delay, and disarrangement of plans caused by slides during excavation, and above all the serious consequences of interruption to operation of the finished canal, are f actors on the side of prevention.
The extensive unloading work done in 1911-1913 failed to stop great slides which were initiated long before. If, after the first conspicuous evidence of weakness of the high banks presented by the break of 1907, this matter of the limiting slopes for a material of given crushing strength had been scientifically investigated and the material excavated to conform to the slopes thus indicated, it is probable that some of the slides could have been prevented, the extension of certain slides decreased, the total quantity of material to be removed lessened, and the time and expense necessary for completing the canal substantially reduced, notwithstanding the greater rapidity and lower cost claimed for dredging as compared with steamshovel work.15
There are now few areas where the process of unloading as a precautionary measure need be applied. One such area will be mentioned later on. Others may be discovered if the conditions along the canal are studied with sufficient care.
Exclusion of rain water from the Culebra slides.-In order to stop the slides it has been proposed to smooth the slopes rising from the canal and cover them with some material impervious to water, such as oil, tar, bitumen, or asphaltum. Cement also has been suggested, either laid as a continuous concrete sheet, or mixed with the surface material by special devices.
15 The undersigned approve the report of the majority of the committee as let forth above, with the exception of the following sentences:
"A thorough study of the problem would give the basis for at least an approximate solution. * *
"If, after the first conspicuous evidence of weakness of the high banks presented by the break of 1907, this matter of the limiting slopes for a material of given crushing strength had been scientifically investigated, and the material excavated to conform to the slopes thus indicated, it is probable that some of the slides could have been prevented, the extension of certain slides decreased, the total quantity of material to e removed lessened, and the time and expense necessary for completing the canal substantially reduced, notwithstanding the greater rapidity and lower cost claimed for dredging as compared with steam-shovel work."
The statements made in these two sentences are, in our opinion, much too strong, and should be modified to read as follows:
"A thorough study of the problem might furnish a basis for a better solution. * *
"If, after the first conspicuous evidence of weakness of the high banks presented by the break of 1907, this matter of the limiting slopes for the actual material in this area had been scientifically investigated, and the material excavated to conform to the slopes thus indicated, it is possible that some of the slides might have been prevented, the extension of certain slides decreased, and the total quantity of material to be removed lessened. On account of the greater rapidity and smaller unit cst of dredging as compared with steam-shovel work, it is doubtful whether the time and expense necessary for completing the canal could have been substantially reduced, but it is possible that the unfortunate closure of the canal in 1915 might thus have been prevented."
The problem of preventing slides by fixing limiting slopes is not in our opinion so simple as to justify the positive statements in the majority report, for which substitutes are here offered. We believe that the crushing strength of the weakest portions of the heterogeneous mass involved in this problem could be determined in advance very roughly only. The shearing strength is still more important, as is also the direction and extent of any lines or planes of weakness. Many other elements difficult of determination help to complicate the problem and vitiate the result of any theoretical solution.
We agree with those who have signed the majority report that a thorough, scientific investigation of the slides from 1907 to date would have been fully justified by the possibility of reducing the time and cst -necessary for completing the canal and by the certainty of operation after completion. It would also have developed information which would be a valuable aid in great excavation projects in the future. We believe that further intensive studies even at this late date are fully justified by similar reasons. They may serve to furnish further engineering information for future use elsewhere and possibly reduce the cost of maintaining the canal.
ARTHUR P. DAVIS.
JOHN F. HAYFORD.
HENRY L. ABBOT.
ACADEMY OF SCIEMNCES.] XVIII.]I
REPORT OF THE COMMITTEE.
The proposers of this scheme f ail to take into account several important factors. No surface covering of the slides can remain intact because of the irregular, fitful movement of the ground. Rainwater doubtless accelerates the advance in some degree but the main movement is of deep origin not immediately affected by surface water. If the surface cover could be maintained, which the canal engineers consider very doubtful, the movement would go on, for the ground-water level could not possibly sink (if, indeed, it were lowered at all) to the depth where the deformation is greatest.
From the practical standpoint it is probably desirable to have the movement of the Culebra slides continue until all material which must reach the canal has been removed, provided its advance is not too rapid. The dredges are reported in December, 1916, as able to keep the channel clear and remove the slide material as it comes in. The slides are growing smaller by removal and must in no great time cease to threaten the canal.
CONCLUSIONS AND RECOMMENDATIONS.
In view of the character of the slides in Gaillard Cut and the conditions prevailing there at this time, as set forth in preceding pages, the committee submits certain conclusions and recommendations bearing on the future security of the canal. The committee regrets that at this late stage in the development of the slides it can not offer more effective suggestions for the control of the slides which closed the canal in September, 1915, and are still active. Although no such large slides in the Cucaracha formation can possibly occur again at any point' along Gaillard Cut, the banks of the Canal have not reached a state of final equilibrium in all places and the committee has considered it wise to point out certain steps which it is believed will assist in protecting the canal from future trouble, and lead to a justifiable feeling of
secuity.ACTIVE SLIDES.
.The East and West Culebra slides are the only ones active in December,' 1916, as far as the committee knows. The phenomena and history of those slides, the character of the rock involved and the nature of the movement, all show that the slides will not stop until a certain undetermined portion of the material has been removed.
The rate of future movement of these slides will fluctuate through dependence on changing internal conditions. At the present time, with the canal between the slides widened to 500 feet, the dredges are able to keep pace with the advance without interfering with navigation. Only a decided -acceleration of the movement can again block the channel and the chances of such a situation arising decrease as the mass of the remaining slide diminishes.
Rain water, which gains access to the moving mass by all kinds of fractures, may be an important agent in accelerating the flow as it works its way downward. In the preliminary report we recommended the diversion of all drainage from surfaces adjacent to these slides, from fissures bordering the slides, and the establishment of a system of open drains on the slides themselves. Mr. Comber reports that considerable work of this character was done on both sides of the canal in the first half of 1916.16 The committee regards this matter as of importance and renews its recommendation that surface water be diverted or carried away from the outer cracks and the moving slides by all practicable means. When a condition of inactivity appears to have been reached its permanence will be rendered more probable by the diversion of surface water.
INACTIVE SLIDES.
General concluions.-All slides now referred to as "dead," "quiescent," or "inactive," the renewal of which would affect the operation of the canal, deserve continued observation. Movement may have been simply arrested and apparent stability may depend on a delicate balance subject to disturbance at any time. The original surfaces of slipping and subsequent fractures in the slide mass are avenues for attack by percolating water. Renewed movement
16 Isth. Canal corn. Ann. Kept. 1915-16, p. 305.
17
18 PANAMA CANAL SLIDES. [MEO VSNTOL
in some slides is probable. The amount and character of material involved, and its position relative to the canal, determine the importance of each such slide from the engineering standpoint. It is desirable to emphasize that the development of the weakness resulting in recurrence of a slide may be a slow process. The Buena Vista slide, which began in 1908, was inactive for some years, and renewed movement in 1915 is a case in point.
Cucaracha slide.-The committee believes that the present inactivity of Cucaracha slide is mainly due to the restraining influence of intrusive bodies of basalt. (See p. 57.), The number of these masses is not known-presumably there are some in addition to those revealed in present exposures. The strength of these natural supports is' also unknown. They may prevent further sliding indefinitely. Some of them may give way, as did the "dike" of 1913.
If sliding is renewed, the volume is likely to be much less than in the great slide of 1913-14, and the operation of the canal may not be interrupted. There can be no future slide at Cucaracha comparable in volume with either of the Culebra slides.
The only protective measure we can recommend is the establishment and maintenance of the most thorough surface drainage system practicable under current conditions. If circumstances permit the use of the hydraulic grader for reduction of load, such work is highly desirable.
POSSIBLE EXTENSIONS OF EXISTING SLIDES.
The great slides of the Culebra district, which are wholly within the Cucaracha formation-excluding the small portions from basalt and breccia masses-can not possibly be greatly enlarged within the area occupied by it because there is but little of the Cucaracha rock available for extensions. We wish to emphasize this point, that all possible additions to these slides are small in comparison to the masses' now in motion, and of very little importance when compared with the total amount of slide material already removed. The conditions of the rock adjacent to each of the large slides will now be considered.
West Culebra slide.-At the rear of this slide the basalt intrusion of Zion Hill and the breccia mass of Culebra Hill and its northern shoulder, called Hodges Hill,.place a distinct boundary. Jointed masses may slough off from the broken faces of these hills but to a limited extent only.
On the north the line descends to lower ground, in touch with Hodges Hill slide, which has been removed to a large extent. A large extension within the limits of the Cucaracha f ormation is not possible.
Between the slide and Contractors Hill, on the south, lies the only considerable mass of undisturbed Cucaracha formation adjacent to the canal in the whole Culebra District. The committee believes that this mass has not yet suffered the initial deformation in depth essential to the Culebra type of slide. It may not yield if let alone, but the committee believes that the factor of safety should be increased by further removal of the load. Some shovel work was done here in 1911-13.17
The area in question is roughly rectangular in shape and embraces about 11 acres. It extends for about 800 feet north from the face of Contractors Hill and 600 feet back from the canal, to a low ridge running out from Contractors Hill. The north limit is a part of the present south boundary of the active slide. The slope rises abruptly from the canal bottom for 150 feet, to the lower terrace made by former shovel work. The terraced slope rises gently 50 feet, and then comes the ridge whose crest is at 280 feet above sea level. The situation and topographic form of this tract may be seen in plates 18 and 39. In plate 24 the fracture now bounding the West Culebra slide on the south is shown, the terraced slope on the left of it belonging to the area under discussion.
The southern zone of the West Culebra slide has not as yet settled or moved toward the canal as much as other parts. We are uncertain how much effect the future movement of the slide may have on the stability of this adjacent mass of similar and, as yet, undisturbed Cuca171 doubt the wisdom of removing the materials mentioned. It was the slipping out and removal of such materials on the north that caused or facilitated the breaking off of the overhanging edges of that hill. I think it should be left as it is-I. C. BRANNEEI.
ACDEY FScE1~ .] REPORT OF THE COMMITTEE. 19
racha rock. Some sliding took place in this area in the years 1907-1911, when the cut was much shallower than at present, but all of the disturbed material was removed. At the contact with the breccia mass of Contractors Hill there was a small open crack, in December, 1915. There was no indication of recent movement on that contact but the committee in its preliminary report recommended the filling of the crack, to prevent access of rain water. The crevice has now been sealed and a drain constructed to divert water from the contact. 18
The committee recommends that the crest of the ridge from Contractors Hill be removed, at least to the 250-foot level, and that the steep bank of this area facing the canal opposite Gold Hill be reduced to a slope of about 1 on 3.5. -The hydraulic grader would seem available for the latter task and shovels could be put at work on the crest of the ridge, for the tracks used in earlier work passed through the saddle at the end of the ridge, at the 250-foot level. The hydraulic grader could also reduce the bank toward the West Culebra slide.
East Culebra. slide.-The Cucaracha rock forms the divide of gentle slopes on the eastern border of this slide, seen in plate 1. Additions to the slide from this side are possible, but because of the distance from the canal and the consequent low slope on which movement must take place, if an extension should occur such possible additions would not be large. Gold Hill limits the slide to the south.
Adjacent to the East Culebra slide on the north the greater part of the canal bank for a distance of 2400 feet has been already reduced in slope by slides, the bases of which are within the Culebra formation. Hagan's slide, in the north half of this area, exhibited the characteristic features of the Culebra slides, including upheaval of the floor of the excavation. (See pls. 43 to 46.) A point of undisturbed material immediately south of Hagan's slide, which rises abruptly to the 150-foot level, should be graded down and other high ground nearer the East Culebra line should also be reduced. Portions of the bank adjacent to the East Culebra slide have already been reduced by the hydraulic grader and we recommend the extension of this work to all the higher ground as far north as Hagan's slide, and for 500 or 600 feet back from the canal.
Cucaracha slide.-So long as the Cucaracha slide remains inactive additions to its area are possible only at the southern limit and these are important only as they may affect the stability of the greater, inactive mass. Should the large slide resume activity, it could receive only slight increments from the basalt of Gold Hill on the north, while the large basalt masses on the east will prevent extensions of much importance on that side. No further reduction of pressure in the upper part of this slide is practicable.
STABILITY OF THE HILLS OF CULEBRA DISTRICT. It has been stated by some persons unfamiliar with the geology of the Culebra district that Gold Hill and other hills near the canal are unstable and that the canal will be unsafe until those hills have been "razed." The geological conditions, which are described in Appendix B and in the discussion of the Culebra district (p. 47), give strong evidence, on the contrary, that those hills are, in the main, self-supporting masses of hard rocks whose foundations are far below sea level. The excavation of the canal and borings in its bottom show that between Gold and Contractors Hills there is a narrow belt of sof t Cucaracha beds; yet to the present time there has been no upheaval of the bottom of this part of the canal, nor any other sign to indicate that the hills have settled. The presence of joint fissures and other fractures may lead to the sloughing off from time to time of masses, here and there, most of which will not affect the canal, but some of which might result in very temporary partial obstruction.
It is believed that the canal-ward face of Gold Hill will not be subject to rock slips except possibly at the sharp angles on the canal bank where the East Culebra and Cucaracha slides may have caused fractures where small masses might in time become detached.
The cliff of Contractors Hill facing the canal has been cut into the breccia mass and shows no weakness. Where the steeply dipping fault plane turns from the canal bank to the north
18 Isth. Canal comm. Ann. Rept., 1915-16, p. 305. Report of W. G. Comber.
20 PANAMA CANAL SLIDES. [ VWrISNTONL
face of the hill (see pl. 48) the breccia overhangs somewhat. It is desirable to know definitely whether or not this fault plane at this place continues for some depth at the angle exhibited at the surface. To this end the committee recommends that core borings should be made from the berm by the canal, with a drill giving solid cores, and penetrating the breccia beyond the fault. Some of the holes should be horizontal and others inclined at 300 to 40, and it is desirable to have the precise location and number of these holes determined by a geologist.
The small amounts of additional material which may break off from the faces of Zion and Culebra Hills will fall on the outer zone of the West Culebra slide. The weight of this material can have but slight effect on the slide and it is not likely that material from either hill will ever reach the canal.
FUTURE OBSERVATION AND INVESTIGATION.
importance of further study.-The committee has already expressed the belief that the banks of Gaillard Cut beyond the limits of the active slides have not everywhere reached a condition of permanent equilibrium. The preceding recommendations imply that observation and study of the canal banks should be kept up indefinitely. But this is only a part of the work which the committee believes should be maintained until the slopes adjacent to the canal have reached a condition of unquestioned stability.
In its consideration of the Panama slides the committee has been impressed by the fact that the problems presented by these phenomena have not been investigated until it was too late to prevent or effectually restrain the slides. The committee regrets that the United States engineers in charge of~ the canal excavation have not had the benefit, from the outset, of the best available technical advice in regard to the proper slopes for the canal banks, based on a through study of the rocks in the banks, and in regard to the character of the slides. Dr. Howe, who was attached to the canal force under Engineer John F. Stevens, 1906-7, before much progress had been made in excavation, was occupied mainly in the preliminary geological study of the canal route and in special problems which were pressing at that time. Dr. Hayes, the first geologist called upon to examine the Culebra slides, was sent to the canal in 1910, on the request of President Taft, but remained there for a short time only. On his recommendation a geologist (Mr. MacDonald) was attached to the canal corps and for three years rendered valuable assistance,* although the great slides were already past control. He was the first to recognize the general character of the deep-seated deformation which characterizes the Culebra slides and explains the upheaval of the bottom of the cut, which was a feature of some movements.
It is now too late to determine just what the result of prompt investigation of the Culebra slide phenomena might have been if it had been undertaken as soon as the problem was presented in October, 1907; but there is no room for doubt that the best way to meet an unexpected problem in any engineering project is on the basis of the best obtainable advice, secured as early as possible after the problem appears.
A complete discussion of the slides of the Panama Canal from geological, physical, and engineering standpoints would still be of great importance. The committee regards it as the duty of the United States Government to see that the slides of the Panama Canal Zone are made the subject of a thorough investigation and report, as a contribution to science and because of the value such a work would possess for engineers (in either Government or private employ) who may in future be connected with great excavation projects. This report should furnish an adequate discussion of the slides, dealing with the history of their development, their origin, the nature of the various movements, and the engineering problems of the prevention or control of slides and removal of slide material.
The future work recommended by the committee is desirable as an aid to the preparation of such a report and for its bearing on future plans for safeguarding the canal. Some of the lines of observation and investigation which should be taken up will now be considered.
Underground water and related data.-As underground water is of great importance in promoting slides in various ways, it should be carefully studied in the Canal Zone. These studies should be carried on in undisturbed areas of each of the formations which have been
ACADEMlY OF SCIENCES.] REOTOFTECMMTE.21 XVIII.] EOTO HECM ITE
involved in important slides, especially the Cucaracha, Culebra, and Las Cascadas formations. A fully satisfactory investigation of this subject would require work in the following directions:
1. Determination of water-table in various rocks in the wet and dry seasons.
In connection with other investigations mentioned below, the water-table data might indicate places where drainage was practicable and desirable.
2. Rate of transmission of water through various rocks under different pressures.
This would supplement the studies of water-table and porosity.
3. Percentages of porosity of the formations of slide areas: (a) Absolute; (b) with regard to size of pores.
The percolation of water depends not only on the absolute porosity but also on the distribution of that porosity. A rock mass which has relatively small porosity may have few openings but these may be continuous and relatively large; such rocks may readily transmit water. At the other extreme are rocks the pores of which are very small and which have no noticeable openings. Rocks of this class may have very high porosity and yet have little capacity for transmitting water. The determination of the permeability of the Cucaracha formation, by the work of Mead and MacDonald, has been of value as giving a desirable basis for opinions that underground drainage of that formation is impracticable. Other formations involved in slides doubtless possess different degrees of permeability.
4. The effect of tropical vegetation with the accompanying humus, its removal, and its restoration, on (a) the amount of water which sinks underground in sliding and in undisturbed areas; (b) chemical action, such as oxidation, hydration, carbonation, etc.; (c) the composition of the water; (d) changes in underground temperatures.
There is difference of opinion with regard to. the effect of vegetation on the percolation of water in underlying rocks. A careful study of the effects of the tropical vegetation from the points of view enumerated will lead to knowledge which may be applied in the future to the problem of using vegetation as a protective medium for weak rocks.
5. The nature of the changes which result in the disintegration and decomposition of the various rocks of the Culebra district when exposed to weathering agencies: (a) With regard to volume; (b) with regard to chemical changes, i. e., oxidation, hydration, carbonation, action of acids, etc.
Whether or not ground will slide depends, among other things, upon its mineral texture and upon its fineness of grain as well as upon the amount of contained water. It may well be that studies of the effects of weathering agencies will develop facts which will be helpful in handling the problem of the slides.
At the time the committee visited the canal none of the above suggested studies had been made except some experiments in the use of vegetation.
At that time it was arranged with Maj. Gen. Goethals that Warren J. Mead and Donald MacDonald should undertake these investigations. Their results, so far as they relate to the pore spaces of the Cucaracha formation, have been of great value in the interpretation of the phenomena of the Culebra slides and in showing the impracticability of draining the Cucaracha rock. But the broader inv estigations of the underground water, begun by Messrs. Mead and MacDonald, have been suspended by the action of the canal authorities. The committee regrets this, for further results of importance would undoubtedly have been obtained, though perhaps not of direct application to the active Culebra slides.
MVechanical testing of the roc.-Sliding- is largely dependent upon the strength of the rocks involved, which may in turn be affected by the water content. The strength of the massive igneous rocks is known to be great and need not be determined. However, tests should be made of the strength of the Obispo, Las Cascadas, Culebra, Cucaracha, and other formations of the Gaillard Cut, when saturated with water, when moist, and when dry, and under conditions of rapid and slow deformation. The matter of the strength of the rocks of the Cucaracha type is of importance and merits careful technical investigation.
Detection of movements of earth or rocc.-The repeated examination of threatening or possible slide areas should be supplemented by instrumental observation, at regular intervals, of
80698-24-3
[MEMOIRS NATIONAL
22 PANAMA CANAL SLIDES. [VOL.
well-placed signals kept up for a number of years. By this means initial movement may be detected which might be masked by vegetation or in other ways escape notice.
The reconunendation of the committee in its preliminary report that such work be done has been complied with to some extent and the perfect stability of Gold and Contractors hills for a certain time established.19
These observations would be specially valuable for signals on Purple Hill and other points of intrusive rock in the Cucaracha slide area, and for points in or between inactive slides elsewhere in Gaillard Cut. The importance of early information of pending movements is selevident. It might lead to successful preventive measures and would at least result in the preparation for dealing with a slide so as to reduce to a minimum its possible interference with the operation of the canal.
Earthquake studies.-T here are now two seismographs installed in the Administration Building at Balboa Heights. It would be an advantage if the smaller instrument should be removed to a second station, for instance, Colon, in order that the origin of earthquakes occurring in regions within 200 or 300 miles of the Canal Zone may be more definitely determined. If an instrument could be maintained in Los Santos Province it would aid greatly in the solution of this problem. Some of the stronger shocks felt in the zone have thrown the needles of the delicate seismographs off the paper and left the records incomplete. A low-power instrument, magnifying about four times, would secure a record of the movements of the ground in these cases.
Organization of scientific wor.-To carry out effectively the work of observation and study which has been recommended will require the cordial cooperation of engineer, geologist, and physicist, working on a definite plan.
GENERAL CONCLUSIONS.
It is obvious that the sliding material which enters the canal must be removed; and it is fortunate that the great slides have now progressed so far that they are approaching a condition of equilibrium. Future slides, should they occur in the Culebra district or elsewhere in Gaillard Cut, will be of little importance compared with those that have already occurred.
The studies of the committee, made since the preliminary report was submitted, have not changed its confidence in the success of the canal. It is not unmindful of the labor necessary to deal with the present slides; it also realizes that trouble in the Culebra district may possibly again close the canal for a short time. Nevertheless the committee firmly believes that, after the present difficulties have been overcome, navigation through the Canal is not likely again to be seriously interrupted. There is absolutely no justification for the statement that traffic will be repeatedly interrupted during long periods for years to come. The canal is a success and will serve the great purpose for which it was constructed.
CHARLES R. VAN HISE, Chairman.
HENRY L. ABBOT.
G. F. BECKER.
J. C. BRANNER..
WHITMAN CROSS.
J. F. HAYFORD.
HARRY FIELDING RED.
R. S. WOODWARD.
R. C. CARPENTER.
*A. P. DAVIS.
November 27, 1917. J. R. FREEMAN.
19 Ann. Rept. Governor of the Panama Canal, fisca year ended June 30, 1916, p. 36.
The three members of the committee whose names are preceded hy a (*) have suggested a modification of certain statements in the report. Their suggestion is given on page 16 in connection with the statements in question.
ACADEMY OF SCIENCES.]
XVIII.]I
Appendix A.
HISTORICAL SKETCH OF THE LANDSLIDES OF GAILLARD CUT.
By WHIITMAN CROSS.
CONTENTS.
Page.
Introduction----------------------------------------------------------------------------7....... 24
Slides during the French r6gime------------------------------------------------------------------- 25
Slides of 1904-1907------------------------------------------------------------------------------ 26
Cucaracha slide ----------------------------------- ----------------------------------------- 26
East and West Culebra slides ----------------------------------------------------------------- 26
Slides of 1908 and 1909 -------------------------------------------------------------------------- 27
Cucaracha slide----------------------------------------------------------------------------- 27
West Culebra slide--------------------------------------------------------------............. 27
East Culebra slide-------------------------------------------------------------------------- 28
Other slides-----------------------------.........------------------------------------------- 28
S lides oof 911910 .... .......... .......... ......... .......... .......... .......... .......... ......... 29
Cucaracha slide ----------------------------------------------------------------------------- 29
West Culebra slide- ------------------------------------------------------------------------- 30
E ast CC ueebra lislide .... ............ ........... ............ ............ ............ ........... 31
Other slides-------------------------------------------------------------------------------- 31
Slides of 1911 and 1912--------------------------------------------------------------------------- 31
General statement ------------------------ ------------------------------------------------- 31
Cucaracha slide----------------------------------------------------------------------------- 32
West Culebra slide-------------------------------------------------------------------------- 32
East Culebra slide-------------------------------------------------------------------------- 33
Other slides-------------------------------------------------------------------------------- 34
Slides of 1913 ---------------------------------------------------------------------------------- 35
General statement-------------------------------------------------------------------------- 35
Cucaracha slide----------------------------------------------------------------------------- 35
West Culebra slide-------------------------------------------------------------------------- 36
East Culebra slide .- -- ---------------------------------------------------------------------- 37
Hagan's slide ----------------------------------------------------------------------------- 37
Other slides-------------------------------------------------------------------------------- 37
Slides of 1914----------------------------------------------------------------------------------- 38
General statement--------------------------------------------------------------------------- 38
Cucaracha slide----------------------------------------------------------------------------- 38
West Culebra slide-------------------------------------------------------------------------- 38
East Culebra slide--------------------------------------------- ----------------------------- 39
Slides of 1915 ---------------------------------------------------------------------------------- 39
Cucaracha slide----------------------------------------------------------------------------- 39
West Culebra slide-------------------------------------------------------------------------- 39
East Culebra slide-------------------------------------------------------------------------- 40
Slides o of 11916---.... -........ .-I-.... ........ ........ ........ ........ ........ ........ ........ ..... 41
General statement-------------------------------------------------------------------------- 41
West Culebra slide-------------------------------------------------------------------------- 41
East Culebra slide ---------------------------------------------------------- -------------- 42
Cucaracha slide.------ --------------------------------------- --------------------------------- 42
B uena VV ista lislide .. ........... .......... ........... .......... ........... .......... .......... 43
Review------------------------------------------------------------------------------43......... 4
23
24 PANAMA CANAL SLIDES-CROSS. [MEvmOHS NATIONAL
INTRODUCTION.
The earliest reported landslides along the route of the Panama Canal occurred in 1884,, soon after the commencement of excavation by the first French company. The "slides," as they are commonly called, for short, increased in number and extent with the progress of the work until the desired limit of excavation had been reached, after which they decreased materially, except in the deepest part of the cut, in what is called in this report the Culebra district.
A complete description of these earth and rock movements, with a discussion of attendant conditions and the character of the phenomena, would be of great scientific interest and of much practical value to the engineering profession. Unfortunately the data for a satisfactory account of the slides do not exist. The movements have been observed principally by engineers, whose measure of their importance, under the dominating ambition to accomplish a gigantic task in an allotted time, has been the degree of inconvenience, delay, and expense caused by the slides. There are, in consequence, few records of the nature of the slides or descriptive notes of their progress available for a proper scientific study.
The present sketch makes no pretension to completeness or great accuracy. It has been prepared for the use of the committee, as one essential to a full understanding of the problem in hand. In the absence of any other connected account of the slides it has been considered desirable to print it as an aid to the reader of the committee report.
Referring for greater detail to the appendixes of this report dealing with the geology of the Canal Zone and the cause and physical features of the slides, it may be well to recapitulate in this place, before considering the course of the slides, some of the principal underlying conditions. The Canal Zone is a region characterized by many prominent hills of hard rock separated by valleys and intermediate smooth slopes underlain by variably softer rocks. The principal hills are made up, so far as their structure is known, of intrusive basalt or of a massive, indurated, basaltic breccia. The breccia of some hills is believed by Mr. MacDonald to be intrusive in its present position, having been forced up through softer formations by ascending basalt lava. These intrusive masses are nearly vertical in position and are in large degree self-supporting columns, since their bases are far below the surface. An element of weakness arises from the fact that the masses are much jointed, as seen in Gold and Zion Hills, they are traversed by some variably inclined major fissures or fault planes, and, as has been shown in Gold and Contractors Hills there was a bulging or flaring out of these masses at least near the surface existing before the canal was cut, expressed with some exaggeration by the current term "mushroom form." Many fracture planes passing through the projecting parts might be so situated as to detach portions of these masses and cause them to rest entirely on the softer intruded rocks.
In the somewhat lower country north and south of the Culebra district there is an alternation of variably hard and soft formations of no great thickness. In the Culebra district a single, nearly homogeneous, and exceptionally weak formation extended from the surface to below the bottom of the proposed cut.
The whole Canal Zone is traversed by a network of faults of various dips and spacings. Most of these have been revealed by the canal cut. The position of these faults in relation to the canal is an element of strength or weakness as the case may be.
The slides represent a perfectly natural adjustment following more or less closely on the rapid cutting of a deep trench through rocks of various weaknesses inherent in their constitutions or in the conditions of their occurrence.
The most important published sources of information concerning the Panama slides which have been utilized in preparing this account are the following:
1. Annual reports of the Isthmian Canal Commission and the Panama Canal for fiscal years from 1905 to 1916. These contain reports by the chief engineer and division engineer.
2. Monthly reports of the chief engineer contained in the Canal Record, Volumes I to IX, 1908-1916.
3. Unsigned notes and articles in the Canal Record, which has been issued weekly under the authority and supervision of the Isthmian Canal Commision since September 4, 1907.
4. Slides at Panama by Maj. Gen. George W. Goethals, forming a Supplement to the Canal Record of January 5, 1916. 17 pages; illustrated.
ACADEMY OF, SCIENCES.) ITRCLSEC FLNSIE FGILR U. 2 XVIII.]HITRCLSEC OFLNSIE OFGILR CU.2
5. The Dry Excavation of the Panama Canal, by Maj. Gen. George W. Goethals, Transactions of the International, Engineering Congress, held in San Francisco, September 20-25, 1915. Paper 10, in Volume I, pages 335-386, San Francisco, 1916.
6. Report of Ernest Howe, geologist to the commission in 1905-6. Annual Report Isthmian Canal Commission, 1907, pages 108-138. I have also had the benefit of personal conference with Mr. Howe.
7. Report of C. Willard Hayes, geologit, Canal Record, Volume IV, page 115.
8. Reports of Donald F. MacDonald, geologist, in the annual reports for 1912, 1913, and 1914, and various articles in the Canal Record, Volume V and VI. Bulletin 86, Bureau of Mines, 1915, "Some Engineering Problems of the Panama Canal in their Relation to Geology and Topography." Also several other papers, in periodicals.
Much important information has been derived from the official maps and the extensive series of excellent official photographs on file in the office of the Canal Commission at Balboa.
SLIDES DURING THE FRENCH REGIME.
Soon after excavation for the canal was begun in the region south of Gold Hill, local slips or slumps of surface materials into the newly made cut took place. All references to these early slides agree in representing them as mud flows of superficial material. These slides occurred in a small drainage basin between Gold Hill and the old village of Cucaracha. The large Cucaracha slide area of the present time probably includes all of the territory within which the early movements near Cucaracha occurred.
The slides in the Cucaracha region took place principally in the rainy season and involved apparently only the superficial zone of dark red, impure clay soil, etc., 20 to 30 feet in thickness. Where the east edge of the canal cut through this material the natural slope of the ground was toward the canal. Beneath the surface materials was a relatively impermeable gray clay rocknow called the Cucaracha formation-and the plane between these two materials is reported to have been the general surface on which the sliding took place. It is probable that during the dry season the ground-water level sank about to the upper surface of the Cucaracha beds. On return of the rainy season the surface soil, etc., became saturated and the contact clays soft and slippery. Small slides were a natural result.
The French geologists, Bertrand and Zurcher, who were attached to the second French commission, made a study of the conditions in the Cucaracha district in 1898, and characterize the material principally involved' in the local slides as a residual, impure clay.
It did not escape their attention that the underlying rock-the Cucaracha formation-was peculiarly weak and yielded to pressure very easily when saturated with water. They do not refer to slides known at the time of their visit as due to breaking down of the rock at any considerable depth.'
The early slides of the Cucaracha district took the form of mud flows descending over the steep slopes of the cut and spreading out on the terraces. The motion of such a mass is apt to be rapid at first, but it soon becomes slow and resembles that of a glacier. Slides of this type are soon arrested if the material of-'the advance portion, or toe; is able to accumulate, as on a bench, for such accumulation serves as a buttress to hold back the rear portion. On the other hand, the removal of the toe of a slide from a working level in the canal prism, necessary to maintain railroad tracks, tended to keep the slide in motion.
A good illustration of the condition of the east bank of the canal from Gold Hill southward as it was in 1890, just after the cessation of work by the first French commission, is given in plate 47 of the Annual Report of the Isthmian Canal Commission for 1914. All of the bank shown in this view has been involved in the later Cucaracha slides.
There are no accessible records to show how much these slides interfered with the work of excavation under the French companies. No doubt they were much more troublesome then than were slides of similar extent in later years, when improved machinery and methods of work had greatly facilitated the removal of such material.
I Marcel Bertrand and Phillipe Zurcher, Ittude g~ologique sur 1'sthme de Panama. Rapport de la Commission. campagnie Nouvelle du canal de Panama, etc. Annex pp. 85-120. Paris, 1899.
PANAMA CANAL SLIDES-CROSS.
[Ms LNIWRrs NATIONAL
[VOL.
It appears that the slides were active in the period from 1885 to 1889, during which the first French company was at work. From 1889, when this company failed, to 1905, these slides were inactive or of minor extent, mainly for the reason that little or no work was done in the Cucaracha district by the second French company.
SLIDES OF 1904-1907.
The United States took possession of the Canal Zone on May 8, 1904. For some years before that time there had been no new excavation work; the banks of the French cut had become overgrown with jungle vegetation, and such Small slides as may have occurred during this period have not been recorded. Excavation of notable amount did not begin under American engineers until 1905. With the resumption of active work, the conditions most favorable for slides were again realized, and as work progressed they became more and more important.
Cucaracha slide.-.After work was resumed in the Cucaracha area, in July, 1905, small slides occurred from time to time, especially during the rainy seasons. Like the earlier movements, these consisted chiefly of surface material and were of small extent. In 1906 some of the slides involved the spoils of French dumps situated near the cut.
For two years or more the slides were not large enough to interfere very greatly with the work of excavation. In the dry seasons it was the practice to cut a wide berm outside the canal prism, upon which anticipated slides could accumulate without overflowing the inner working levels.
In the wet season of 1907, slides in the Cucaracha area showed increased activity and, filling the outer berm, flowed into the canal prism. Early in October, the slide just north of Cucaracha village suddenly developed, and a large mass flowed over all berms onto the floor of the cut, burying several tracks and catching some steam shovels. By October 9 it was estimated that 500,000 cubic yards were in motion. Plate 5 is a view of this slide taken on October 5.
For two weeks the rate of motion of the slide was maintained at about 14 feet per day. During this time the inflow of material was more rapid than the excavation by all available shovels, working night and day. The rate of flow decreased to 4 feet per day at the end of October, and by the middle of November the slide ceased to interfere with transportation through the cut.
East and West Culebra slides.-The northern part of the Culebra district appears to have been free from slides of any kind, comparable in importance with those of the Cucaracha area, until 1907. At least the minor slumps and mud flows which may have occurred do not seem to have been recorded. The condition of the east bank north of Gold Hill in 1904, soon after the United States took possession, is represented in plate 27.
-In the summer of 1906, Mr. Howe observed a notable crack in the ground in front of the boarding house at Culebra, 100 feet or more back from the western edge of the canal cut. The significance of such fissures in this vicinity was not established at that time, and it is not known whether this crack was connected with the slide of 1907, which is to be described.
Gen. Goethals has given a brief description, which will be quoted in full, of an important slide which occurred in 1907, in the West Culebra areas.
In October, 1907, a crack developed 50 or more feet from the face of the bank on the west side of the canal near the old railroad station at Culebra. Subsequent to the cracking, the ground to the west of it showed perceptible settlement, though the elevation of the portion between the crack and the prism remained unchanged; this settlement was accompanied by a bulging up of the bottom of the canal sufficient to raise the shovel. At the time this occurred. the shovel was at elevation 175, and the subsiding portion of the upper bank was at elevation 275. 'Ultimately the portion between the crack and the prism settled down, sliding into the prism. Subsequently new cracks developed farther back, followed by similar settlements and movements into the cut, involving large masses of material; the same condition developed on the east side.2
It is not clear how much of the movement referred to in the last sentence followed closely on the original slide. No mention of this notable slide appears in the notes of the Canal Record of the time, in the October report of the chief engineer, or in the annual reports of either the division or the chief engineer covering the latter part of 1907. Although Gen. Goethals gives
2 Trans. Int. Eng. Cong, Vol. 1, pp. 372-373. Practically the same description appears in Supplement to Canal record, Jan. 5, 1916, p. 7.
26
ACADIY OF CFNS.HITRCLSEC OFLNSIE OFGILR CU.2 XVIII.] "~n- ITRCLSEC FLNSIE FGILR U.2
this quoted description of the West Culebra slide of 1907 as an introduction to a discussion of the type of slide called "breaks by the engineers, the fundamental difference between this slide and the surface flows of the Cucaracha area-active at the same time-does not seem to have been appreciated for some years.
The West Culebra slide of 1907 was, however, the forerunner of the deep-seated movements on a large scale, which since then have characterized the Culebra district and to which the recent slides obstructing the canal clearly belong. The Cucaracha formation was shown by this slide to be incapable, locally at least, of sustaining its own weight in steep banks 100 feet high. Borings had shown the formation to extend to below sea level and the bottom of the canal opposite this slide was planned to be at 40 feet above sea-level, or 13 5 feet below the foor of the excavation at the time of the slide. Much larger movements of the same character were thus foreshadowed by -the first slide of this kind.
It is not clear from available data as to how much sliding took place in the East Culebra area in 1907, nor what its character may have been. Gen. Goethals says: "In January, 1907, a slide occurred on the east side opposite Culebra," 3 but does not refer to its character. No other reference to slides in this area has been found.
SLIDES OF 1908 AND 1909.
The special feature in the development of the slides during the years 1908 and 1909 was the occurrence of several small local movements, each one exhibiting some peculiarities in either the phenomena of the slide or the underlying conditions. Some of these slides were more than the slipping of surface clay or earth.
Cucaracha slide.-Slides in the Cucaracha area continued to give some trouble and caused interruption to transportation on several occasions, but, on the whole, activity subsided for a time, after the surface material immediately adjacent to the upper berm had slipped into the cut.
The character of the larger slides in this area in the earlier years of the American work is well illustrated by Plate II of the annual report for 1908. This represents a mud flow near the village of Cucaracha in October, 1908, which for 10 days maintained an average advance of 14 feet each 24 hours, the same rate exhibited by the slide of a year before. The material in motion was estimated at 600,000 cubic yards, in an area of 34,445 square yards. The flow passed over a broad, upper bench and covered several tracks on the bottom level.
The area of the slide gradually increased to 27 acres and the annual report For 1907-8 states that 670,017 cubic yards had been removed from this slide prior to June 30 1908, and that 700,000 cubic yards were still in motion at that time. The length of the slid along the cut had increased to 2,700 feet, but sliding was not in progress for this entire dist ce at any one time. The conditions along this stretch of 2,700 feet, from Gold Hill to nea D'ucaracha village, as they were in June, 1909, are clearly shown in plate 34 of the annual repe It for 1909. Between the upper berm and an irregular line of vegetation above it there was a bare, steep slope, partly due to excavation, but largely denuded by the slipping of many sections of superficial red clay.
By the beginning of the dry season, late in 1909, the Cuaracha slide had become inactive and the engineers anticipated no more trouble in that area.
West Culebra slide.-Slides involving material belonging to the Cucaracha formation developed on an increasing scale during the yeais 1908-9, but, up to the latter part of 1909, they were scarcely mentioned in the reports of the engineers or in the news items of the Canal Record. After some unusual movements in November of 1909, the Record gives a brief review from which the following extracts are taken:
For about two years there has been a crack in the west bank at Culebra, nearly 2,000 feet long and extending back toward the village a distance of from 600 to 900 feet from the center line. There has been no appreciable widening of this break as a result of the recent rains, but it is considered as defining the limits of the Culebra slide. It is estimated that 2,000,000 cubic yards is comprehended within these limits.***
3 Supplement Canal record, January, 1916, p. 11.
P-.ANAMA CANAL SLIDES-CROSS.
[MfEMOIRS NATIONAL
[VOL.
There have been three distinct movements of the Culebra slide at Culebra village since the heavy rains of November saturated the ground. The first of these occurred on November 16, when a mass of earth estimated at approximately 110,000 cubic yards broke away from the west bank and moved into the prism of the canal, carrying a portion of the old main line of the Panama Railroad and putting three construction trains out of service. On November 28, this material moved farther toward the center line and filled in the pioneer shovel cut. * During the morning of December 4 the west bank gave way a few hundred feet north of the first slide and settled down into the cut. The earth raised a mound 30 feet high in the prism of the canal * Although these movements * are in a sense local, they really form parts of one large movement of material ** toward the prism of the canal.4
From these statements it appears that, although the slide material is referred to as "earth," the banks of the canal composed of the soft Cucaracha formation had begun to break down and slide. The phenomena of these slides were not at the time interpreted as evidence of deepseated crushing of the rock and as indicating a huge slide like that of October, 1907, but in the light of later slides it becomes plain that the great deformation of rock in the West Culebra slide area was in progress.
The amount of this movement as distinguished from slips of surface material, is perhaps best indicated by the Annual Report of Col. Gaillard for the fiscal year 1909-10. Earlier reports by Col. Gaillard had referred to all the slides as of superficial character but in this report "breaks" are recognized and the following statement is made. "The largest break occurred on the west bank of the canal at the town of Culebra. The total broken area covered l01 acres. Prior to July 1, 1909, 675,634 cubic yards of material had been removed and 1,005,388 cubic yards during the fiscal year ending June 30, 1910."1 No doubt an important part of the latter amount belonged to 1909.
The appearance of.a small slide in October, 1909, is represented on plate 15.
Toward the end of the dry season of 1908, early in April, a portion of a French dump on which the village of New Culebra had been built started to slide, undermining a number of houses. This movement extended 170 feet back from the edge of the cut and involved at one time 40,000 cubic yards. The unconsolidated dump d6bris was dry and it slipped on a gently sloping surface of harder material which was "smooth and soapy to the touch." 6 This dump caused trouble at intervals after this slide until it was completely removed.
East Culebra slide.-There appear to have been few slides in the East Culebra area during 1908 or 1909. In view of the extensive slide of 1915, it is noteworthy that the development of deep-seated weakness in the rocks did not proceed as rapidly as in the West Culebra area.
Althoi gh the annual report of Col. Gaillard for the fiscal year ending June 30, 1909, does not specifi illy refer to any important slide in the East Culebra area it appears that a slide of the type 1: er called "break" began on the east bank opposite Culebra in the first half of 1909. This slide first mentioned in the annual report for the fiscal year 1910, page 155, as the second largest "b ak" along the canal and there it is said that "prior to July 1, 1909, 166,018 cubic yards of n terial were removed from this break." Probably this slide developed slowly during the latter half of 1909, for no direct reference to its movement during that period has been found.
Other slides-A minor slide, destined to cause considerable annoyance for some years, began in February, 1908, on the east bank of the canal, opposite the village of Las Cascadas. This slide began in the dry season, and its material is characterized by Col. Gaillard as "fairly dry." The slide soon extended for 260 feet back from the edge of the cut and in April, 1908, embraced 100,000 cubic yards. According to the Canal Record (vol. 1, p. 250) the material belonged entirely to the superficial coating of earth, 20 to 30 feet thick. There was little if any motion here in the latter part of 1908 or in 1909, although there were small slips near by in the latter year.
On the east bank of the cut, opposite the Whitehouse yards, another small slide began its course in September, 1908, and caused much annoyance by covering railroad tracks. The Powderhouse slide, slightly to the south of the Whitehouse, began its movement in 1909, and these two ultimately joined.
4Canal record, Vol. III, p. 115, Dec. 8, 1909.
5 Ann. nep. Istli. Canal Cor., 1909-10, p. 155.
6 Canal record, Vol. 1, p. 250, 1908.
28
ACADEMY OF SCIENCES.]HSOIA SKTHOLADLDSFGILRDC xviii.] 1HSOIA KTHO ADLDSO ALADCT
29
At Paraiso, at the south end of the Gaillard Cut, the filling of a small valley, the floor of which sloped toward the cut, began to move in March, 1907, but it did not amount to much until April, 1908. Its steady motion for a month or more kept the shovels at work removing the material as it came down. The slide is said to have taken place on the old surface of the valley and it finally extended back some 600 feet from the edge of the cut.
A small slide started in November, 1908, near the old native town of Buena Vista, on the west bank, in the northern part of Gaillard Cut, but it was too small for special mention at that time and deserves notice here only because it was one of' three active slides at the end of 1915. Another small slide, due to the widening of the cut, began its course in September, 1908, at Haut Obispo, on the east bank.
During these two years the general work of the canal was more or less delayed by the slumping of loose materials in railroad fills, the Sosa-Corozal dam,' the dumps of the quarries at Porto Bello and Ancon Hill. These slumps illustrated the phenomena attending the attainment of stability of slope in loose materials under various conditions. The interest attaching to these slumps as bearing on the slides of Gaillard Cut lies in the character of the movement where loose material rests on a soft, yielding base. The upthrust of material at the toe of such slumps was later thought to resemble the upheaval in the floor of the canal cut opposite certain of the slides, and already exemplified by the slide of October, 1907, in the West Culebra area.
SLIDES OF 1910.
Accompanying notable progress in excavation, there was a marked increase in area and volume of the slides of Gaillard Cut in 1910. The superficial slides continued in the Cucaracha area, but by the end of the year they had become almost inactive.
The principal development of slide action was in the East and West Culebra areas and in the initiation of smaller, local slides in other parts of the cut. Those of the Culebra district became more and more clearly characterized as involving the rupture and crushing of rocks at some depth. The phenomena of upheaval in the bed of the cut contemporaneous with subsidence in some adjacent but not immediately adjoining areas became not uncommon. These slides and others due to the slipping of rock en masse on some fracture plane led the engineers to recognize a form of landslide which they termed "breaks," in contrast to the slides of more superficial character.
An examination of the landslides from the geological standpoint was made in November, by Dr. C. W. Hayes of the United States Geological Survey, at the request of President Taft. His preliminary report was published December 7, 1910.
The increasing disturbance of the work of excavation by the slides in 1910 led to serious consideration of practicable means of restraining them and to the adoption of the plan of decreasing the pressure in the upper slopes adjacent to slide areas through the removal of rock and earth by shovel work.
Cucaracha slide.-During the dry season of the first quarter of the year the movement of this slide lessened so that the shovels were able to clear a berm below it 40 to 100 feet in width and 170 feet above sea level, on which it was hoped that the slide material might accumulate for some time without interfering with transportation on the lower levels in the cut proper. Early in, April the shovels were all withdrawn from this berm.
Reviewing the situation in April, the Canal Record states that "A mass of earth and rock, estimatedl at from 500,000 to 800,000 cubic yards, is apparently slipping on a smooth substratum of rock, the movement being perceptible at all times of the year, but considerably accelerated by the rains from May to December." The magnitude of the slide up to that time is shown by the statement that 1,850,000 cubic yards have been taken out," apparently referring to Cucaracha slide material.7,
With the beginning of the rainy season the anticipated acceleration of the slide occurred, and in the night of May16-17 some 500,000 cubic yards of mingled mud and rock ini the porI Canal Record, Vol. III, p. 250, 1910.
30 PANAMA CANAL SLIDES-CROSS. [MEORSNTONL
tion of the slide adjacent to Gold Hill moved down the east bank and across the botton of the cut to the western side. The Pioneer Cut, in the center, was filled for a distance of 900 feet and all through tracks were covered.8 This movement soon subsided and communication was quickly restored, but a slow slipping of this newly broken mass continued during the summer. Plate 34 accompanying Colonel Gaillard's report for 1910 represents the Pioneer Cut, June 23, as filled by slide debris, for a short distance.
On September 16 a new impulse was given to this mass and again all tracks on the east and one on the west side were covered in a few hours time, and it required more than a week's work to clear them. October 4 a quickened motion of the slide once more caused an invasion of the main cut. On the 22d of this month there was a sudden acceleration in the movement, and the most rapid sliding of a large mass hitherto experienced took place. In 20 minutes the toe of the slide advanced 75 feet, crossing several tracks and filling the Pioneer Cut and drainage ditch. A number of locomotives, cars, and shovels were caught. The material of this slide was a part of the mass which started in May, most of which had come into the cut and been removed.'
It is impossible to ascertain whether the Cucaracha slide up to the end of 1910 contained any material representing the breaking down of the Cucaracha formation. The slides are referred to only as movements of surface soil or dump material, and doubtless they were chiefly, if not exclusively, of the superficial type. But in the excavation of the wide berm beyond the canal prism there must have existed a steep bank of the soft Cucaracha rocks beneath the shallow zone of unstable soil. After the slide of October was removed a wider bench was made south of Gold Hill, and there was exposed in the Cucaracha strata back of it an irregular dike of intrusive basalt extending approximately parallel to the canal cut.
At the end of 1910 the deepening of the canal and the excavation of the safety bench back of it had cut into the Cucaracha formation near Gold Hill, leaving a steep bank about 300 feet high. The absence of slides like those of the West Culebra area may have been due to the buttresslike basalt dike and to other similar bodies of small size, which later excavation has shown to exist in the Cucaracha beds. This protecting influence of the basalt dike was accepted by engineers and geologists alike.10
West Culebra slide.-There is a dearth of descriptive data concerning the West Culebra movements of 1910, but from scattered references to then it is clear that there was a steady extension and unification of the slide resulting from the breaking down of the Cucaracha formation. The actual movement of material during the first six months was not great, but deepening of the cut was in progress and the weakness of the banks surely developing.
The most striking feature in this period was the formation of cracks, with slight contemporaneous subsidence, at higher and higher levels on the slopes of Culebra Hill, endangering buildings and roads. Later in the year many buildings seen in the view of plate 15 were abandoned or removed. The manner in which the upper line of the slide area encroached on slopes occupied by buildings is illustrated by plate 35 in the annual report for 1909-10. The ground occupied by the buildings of this view, taken on June 10, was soon embraced in the slide. In October the development of large cracks led to the removal of several buildings on the upper slopes of Culebra Hill.
The movements in the lower levels of the cut in the latter part of 1910 were accompanied by some upward motion in the floor of the cut, giving evidence of deep-seated disturbance. The combination of developments during the year convinced the engineers of the necessity of reducing the pressure on the slopes by the removal of material from their higher portions. "Two breaks occurred in rather rapid succession at Culebra in the latter part of 1910 * *. The later one of the two determined the change in plan, and in January, 1911, the reduction in the height of the adjacent banks on both sides of Gaillard Cut in the vicinity of Culebra, was directed and begun"'"
'Canl Rcord Vo. IV p.65, 910 11Gen. Goethals, Supplement Canal Record, January, 1916., p. 7.
8 Canal Record, Vol. III, p. 305; 191Q. 9 Canal Record, Vol. IV, p. 65 1910.
ACDEY a'SENE.] HISTORICAL SKETCH OF LANDSLIDES OF GAILLARDCUT. 31
On November 16 a small slide, 16,000 cubic yards, occurred on the north shoulder of Contractors Hill, the material plunging rapidly into the cut. This involved somne massive rock belonging to the projecting "mushroom "-shaped breccia body of Contractors Hill, which may have been detached by joints causing it to rest directly on the soft Cucaracha beds.
East Culebra slide.-The enlargement of the East Culebra slide in 1910 was apparently not so extensive as that on the opposite bank, but it was of the same general character. The slide which began in 1909, opposite Culebra, developed considerably in the first half of 1910. In the annual report 1909-10, page 155, Col. Gaillard speaks of this slide as covering 111 acres, and says that 314,184 cubic yards were removed during the fiscal year." The plates 38 and 39 accompanying his report show the slide to have a steep fracture plane at the rear. An upheaval ridge in the cut opposite is shown in plate 29. The legend of this photograph reads as follows: Bottom of the canal raised for a short distance, through pressure of the broken east bank, 18 feet above its original level." This was apparently a movement like that exhibited during the slide of October, 1907, at a considerably higher level, on the west side of the canal.
Other slides.-The Las Cascadas slide, the initiation of which in 1908 has been mentioned, gave some trouble during 1910, but the volume was so small for most of the year that efforts were made to retard or control the movement by dumping rock and by a line of sheet piling. In December, however, the slide enlarged its bounds and involved rock, as well as surface material. The slide at this time covered several railroad tracks for a few days. These new movements embraced over 300,000 cubic yards and the slide area was increased to 7.6 acres. It is said that the surf ace on which the superficial clays slipped at this time was inclined at 1 vertical to 6 horizontal.1 A view of the slide on December 15 is given on plate 40.
The Whitehouse slide was active at intervals during this year, but did not do much damage.
A slide which took place on May 7, 1910, at La Pita Point, on the steep eastern bank of the canal cutA midway between Empire and Las Cascadas, has served as a type for a group of slides due to structural conditions. According to Mr. MacDonald a fault plane crosses the canal diagonally with a steep northerly dip. It also intersects the east diversion ditch for the Obispo River, and water from that ditch, seeping into the crushed rock along he fault plane, weakened the support for some massive breccia, here resting on agglomerate. As a result a block of rock 150 feet long and extending 175 feet back from the cut slid toward or into the excavation, breaking the diversion ditch and causing its water to flow into the cut for a short time. This mass of rock was also weakened by seepage of water along the breccia-. agglomerate contact.
The rock of the La Pita slide was hard and massive,' as compared with the material of the Culebra slides, and no crushing or deformation of underlying rock took place. The slide (pl. 41) did not flow like those involving the Cucaracha formation nor was it attended by upheaval in the bed of the canal.
A small slide similar in character to the La Pita began in the latter part of 1910, on the east side of the cut, 2 miles north of La Pita. It is called the Bas Obispo slide.
SLIDES OF 1911 ANY) 1912.
General statement.-The most notable features in the landslide development of the period 1911 and 1912 were the enlargement of the East and West Culebra areas by the union of several separate slides and the indication of the ultimate unity of the movements in large areas given by the appearance of deep cracks well back from the border of the excavation. There were some sudden breaks with sharply defined limits, but much of the sliding consisted in gradual advance of sections of the banks on to the catclhment berm, whence the material could be removed without encroachment on the canal prism.
Superficial shallow slides of the Cucaracha type were active at times, bOut decreased in importance, and, as the cut became deeper, south of Gold Hill, local conditions developed which gave rise to the belief that slides of the Culebra type would not occur in that area. In the
"2Canal Record, Vol. IV, p. 187.
32 PANAMA CANAL SLIDES-CROSS. [MVNTOL
parts of Gaillard Cut outside of the Culebra district the floor of the excavation approached the proposed level for the canal bottom. The banks were steep and approximately at full height. This was only 100 to 200 feet, but weaknesses of various kinds developed here and there. In some places sharply defined structural weakness led to block slides, as at La Pita Point, and in others the banks broke down with slides similar to the much larger movements in the Culebra district.
Realization that the serious slides of the deeper parts of the cut were due to the inability of the Cucaracha formation to sustain the pressure of high banks led to the plan of reducing this load by the removal of a great amount of material from the upper slopes, between Culebra and Zion Hills and the edge of the cut and from a corresponding area on the east side of the canal. This work was carried on during 1911, 1912, and nearly to the close of 1913. An apparent decreased slide movement and the delay of development in some fissured areas created the feeling that the lightening of the load was producing the desired effect. But while the lightening was in progress above, the cut was being deepened, the limit of possible disturbance in the rocks was being lowered, and the good effect of the decrease in load was to a considerable extent offset.
At the end of 1912 the floor of the cut in the Culebra district was practically at the level proposed for the bottom of the canal-40 feet above sea level.
Cucaracha slide.-The movements of the Cucaracha slide in the first half of 1911 were unimportant. Early in June the shovels which had been preparing a broad berm at about 95 feet above sea level and extending for 100 feet back from the border of the canal prism ceased work,' and from this time to the end of 1912 the slide was practically in a condition of rest and was confidently assumed to be "dead." For the fiscal year 1911-12 only 170,000 cubic yards were removed from this slide, a comparatively negligible quantity.' In December, 1912, a renewed movement occurred, involving about 50,000 cubic yards, but this revival occasioned no apprehension at the time.
The apparently stable condition of the slope adjacent to Gold Hill in May, 1912, is illustrated by plate 30 of the annual report for 1911-12. The bottom of the cut was then 56 feet above sea level, or 16 feet above the proposed floor of the canal. Above it were four working berms and still higher a steep, bare slope rose to a level perhaps 200 feet above the floor of the excavation. Before the close of the year the lower part of this slope had been still further steepened. The fact that the weak Cucaracha formation did not break down on the steep bank was taken as ground for the belief that the basalt dike uncovered in 1910 would continue to give any needed support.'14 Plate 6 represents one of the intrusives near Gold Hill.
West Culebra slide.-Active development of this slide in 1911 began about the middle of January by a movement of 200,000 to 250,000 cubic yards of material, some of which filled a broad berm at the 135-foot level but did not obstruct the lower cut. The slide occurred opposite the Culebra Hotel and the extension of cracks in the rear endangered and forced the removal of a number of houses. The movement is said to have been anticipated but it came early in the dry season and was followed by other breaks before the wet season began, the time at which renewed movement had been expected. The connected slide area extended at this time for more than a mile along the west bank but not to a uniform distance back from the cut.'5 Evolution of the slide continued steadily during 1911 and 1912, requiring several shovels to remove the material from the catchment berm adjacent to the main cut. The slide seldom reached the lower level of the cut and no direct reference has been found to any upheaval of the floor during 1911.
The encroachments of the slide area upon the upper slopes of Culebra Hill, occupied by buildings, continued, but the cracks signifying the weakening of underlying ground were not closely followed by subsidence, in most cases, so that arrangements could be made in advance for the removal or tearing down and rebuilding of the more important structures, such as the hotel, club house, postoffice, and many dwelling houses.
13 Canal Record, Vol. VI, p. 13, 1912.
14 Gen. Goethals, Supplement Canal Record, January, 1916, p. 6.
1Canal Record, Vol. IV. p. 169, 1911.
XVIII.]
33
Fissures indicating deep-seated def ormation of the Cucaracha beds appeared also on. the slopes of Zion Hill some time before mass movements occcurred. The first appearance of these cracks is not recorded, but in January, 1912, the Canal Record, in reviewing certain phases of the situation near Culebra, says:
Along the edge of Zion Hill, back of the penitentiary, there is a long crack in the earth that is typical of the breaks occurring in this class of slide. In some parts it is from 6 inches to afoot wide. It is apparent tat the toe of the hill has broken away from the main mass, and the depth of the break can only be conjectured; at places it can be sounded to 20 feet below the surface.'6
It is further pointed out that a spring exists on the line of this fissure the water. of which will probably accelerate the movement of the pending slide. Cracks were also known at this time beneath the penitentiary building, between Zion and Culebra Hills.
A slide at Culebra-on-the-dump in February, 1912, is referred to in the Canal Record 17 as a renewal of the slide of two years before, but this new movement clearly' affected rock of the Cucaracha formation. Seven hundred and fifty thousand cubic yards of material started to move at 18 to 24 inches 'a day leaving an abrupt fractured face 30 to 40 feet high at the rear, coming up close to the buildings of the village. The beginning of this slide was not accompanied by upheaval in the front of the cut. The mass gradually broke up as it moved and its condition in June is well shown by official photograph 13-x14, the label of which states that "About 1,000,000 cubic yards of material are moving toward the cut at the rate of 3 feet per day on the bottom slope of about 1 -vertical to 7 horizontal."
.On August 26 a section of the steep canal bank below Culebra, which had for some time formed a projection between portions of the West Culebra slide area, gave way and as it settled there was upheaval in the floor of the cut. Nine hundred thousand cubic yards of material were included in this slide at the outset. It moved rapidly into the canal prism and occasioned much trouble.'8
In the annual report for 1911-12 Col. Gaillard reviews several features of the slide with special reference to those of the West Culebra area. The movement is said to occur where the inclination of the surface is as low as 1 on 10. The slipping of one mass is thought to have taken place on a lignite bed inclined at 1 on 7.
Regarding the relation of slide development to the wet and dry seasons Col. Gaillard remarks that the slide moving on a~ lignite bed-which was probably the one noted as occurring early in 1911 at Culebra-on-the-dump-" developed early in the dry season and has moved at a faster rate for the succeeding four dry months than it has done since the rains have come., The 'breaks' which during the past fiscal year have involved more yardage than the (surface) slides caused as much trouble in the dry season as in the wet season and during the last long dry season of 5 months the proportion of breaks which developed were in excess of those developed during any similar period of the wet season." 1
Col. Gaillard's reference to upheaval in the floor of the cut indicates that such action has been common, though seldom mentioned in the Canal Record. "Most of the slides of the past year (1911-12) were breaks resulting from the failure of an underlying layer of rock of poor quality due to the pressure of the enormous weight which crushes the underlying layer, forces it laterally and causes it to rise up or heave in the bottom of the cut. The heaving at times is 30 feet." He says that the material heaved up is not mainly clay or mud but consists of rock fragments. Where the maximum upheaval of 30 feet took place the sandstone fragments varied from 2 inches to 20 or 30 cubic yards in mass and probably averaged 1 to 3 cubic yards.
East Culebra slide.-The development of the slides during 1911-12 proceeded less rapidly in the East Culebra area than on the west bank, but it was of essentially the same character. With the deepening of the cut the banks gave way here and there, the various slides exhibiting the same phenomena noted in other cases. The breaking up of the slopes did not extend so far from the cut as at Culebra, yet the total acreage recognized as involved was large.
16Canal Record, Vol. V, p. 184, 1912. A map accompanying this review shows relation of the upper slide limit to the the buildings at Culebra.
LCanal Record, Vol. V, p. 197, 1912.
18 The appearance of this slide on September 25 is shown in official photograph 13-x24, at which time several tracks were obstructed.
'9 Ann. Rept. Istbm. Can. com., pp. .161-2, 1911-12.
34 PANAMA CANAL SLIDES-CROSS. [MVNTONL
In his report for 1910-11 Col. Gaillard refers but indirectly to the slides in the East Culebra area aside from a statement of the amount of material involved. In the report for 1911-12 he again mentions this slide area only in general terms. On June 30, 1912, it embraced 15.7 acres and was exceeded in area only by the West Culebra slide. One million nine hundred and sixty thousand cubic yards were removed from it in the fiscal year.
Few accounts of specific slides appear in the records, yet the work of removing sliding material was constantly in progress. The 'details were of less general interest than for the slide adjacent to the town of Culebra.
On the night of February 9, 1911, a section of the east bank of the cut 1,100 feet in length suddenly broke away settling almost vertically about 30 feet and moving laterally displacing the berm at the 135-foot level and all the ground west of it to the Pioneer Drainage Channel which was almost completely closed; 550,000 cubic yards were included in this mass, 375,000 of which were beyond the proposed limits of the canal. All loose material had been removed from the 135foot berm a few days previous to the slide. It occurred in the dry season, and -no blasting had been done in the immediate vicinity for 10 months. Nothing to give warning of this slide had been observed. It serves to show how suddenly breaks in a steep bank may occur while movements affecting areas extending farther back from the bank are commonly foreshadowed by cracks and local settlement. The relation of this slide to Gold Hill and the long stretch of the steep canal bank, now completely broken down, is well shown in plate 41, annual report for 1911-12.
Another slide of small dimensions, occurred on July 4, 1911, adjacent to Gold Hill. This took place suddenly and caught a shovel.
On February 10, 1912, a section of the east bank 500 yards north of Gold Hill broke down covering or disturbing all tracks down to the Pioneer Cut; 250,000 cubic yards were involved at the outset and suoceeding the appearance of cracks south and east of the initial slide much more ground was added to it. Plate 30 represents this slide as it appeared on February 11.
Other slides.-A number of small slides occurred in 1911 and still more in 1912 at various points along the canal cut north of the main Culebra slides. Most of these were on the east side of the excavation. These slides were partly movements of surface material and partly the breaking down of steep banks where some local weakness existed. Most of them do not need special mention.
On the west bank of the cut adjacent to the division office at Empire a slide occurred in September, 1911, which was also similar to the La Pita slide in that it took place on a steeply dipping fault plane, as described by Mr. MacDonald. The initiation of this break was first observed in May, 1910, but notable slipping did not occur at that time. Efforts to retard this slide were made in December, 1911, by blowing away the upper part of the material by dynamite. Eventually the slide embraced 260,000 cubic yards, extending over 2.6 acres, and it became inactive in May, 1913. Owing to the proximity of this mass to the tracks on the west side of the cut it occasioned considerable trouble at times.
On August 20, 1912, there occurred a second and notable break in the east bank of the canal adjacent to the Obispo River diversion. This is known as the north La Pita slide. Like the earlier break this was due to the seepage of water from the diversion ditch into the crushed rock along a sheared fault zone and to the softening of a so-called "mud lava" bed which was highly jointed and underlaid massive breccia. The wall of the cut where the slide occurred rose nearly vertically for about 125 feet. As a result of this land slide "the cut to the north was so completely flooded as to stop all excavation for 3 miles during the remainder of the wet season."
An excellent view of the north La Pita slide is given in plate 42.
In 1911 the adjacent Powder House and Whitehouse slides on the east bank became inactive. These slides combined extend for 4,200 feet along the canal, but in few places do they have a width of more than 300 feet. These have not been given much attention in published accounts of the slides, but they undoubtedly represent the breaking down of the steep bank where local weakness existed and their quiescence indicates an approach at least to a condition of permanent stability.
ACADEMY OF SCIENCES.] HSOIA KTHO ADLDSO ALADCT xviii.] HSOIA KTHO ADLDSO ALADCT
35
In 1912 a number of other small slides which have been active at various times reached a stage of quiet. Among these are the Las Cascadas, which had finally covered 11.5 acres and involved over 500,000 cubic yards of material.
SLIDES OF 1913.
General statement.-At the beginning of 1913 the canal excavation had reached the proposed level for the bottom of the canal, at 40 feet above the sea, except in the Culebra district, where the maximum elevation in the floor of the cut, opposite Culebra, was at 61 feet, early in February. The determined width of the floor had not been so generally obtained. Weakening of the canal banks by deepening of the cut was therefore practically at an end, while reduction of the load on the banks in the East and West Culebra areas, which had been in progress during 1911 and 1912, was continued until December of 1913. The progress in excavation and decline of slide movements was such that early in 1913 tentative plans were made to admit water from Gatun Lake into the cut about October 1, and this step was actually taken on October 10.
The principal slide development of the year was in the Cucaracha area. This slide, considered dead" in 1912, became active in January and soon became a great rock slide comparable with the Culebra slides in size and in many of its features. The slide was active for the remainder of the year and the removal of its material from the cut, together with the lightening of the load above, comprised a large part of the total excavation of 1913.
The Culebra slides decreased in amount, apparently as a result of the lightening of the load above them, and by the end of the yearno further serious trouble from them was anticipated.
Beyond the Culebra district the banks of the nearly completed cut continued to exhibit adjustment to the new conditions. Some new minor slides developed; some moved intermittently; and others became inactive.
Cucaracha slide.-The renewal of landslide movement early in 1913 in the Cucaracha area adjacent to Gold Hill was of special significance because it plainly demonstrated that the apparent stability of a steep artificial slope in soft rocks maintained for a year or two is by no means proof of its permanence, for the reason that the development of an inherent weakness in the rocks of a steep bank to the point of rupture may be a slow process. The intrusive rocks in the Cucaracha beds which were uncovered in 1910 were supposed to possess sufficient strength, as a natural buttress, to hold back the soft rocks of a bank some 300 feet high which might otherwise give way. (See pl. 6.) But the basalt of Zion and Gold Hills and of the smaller intrusions in the area of the Cucaracha slide is very strongly jointed. In the larger hills the jointing is columnar; in other places more irregular. It is probable, therefore, that the basalt of this "dike" was at least irregularly jointed, and the continued pressure of material back of it gradually dislocated, if it did not freshly fracture, the intrusive mass, until it no longer afforded adequate support.
On January 16 and 17 a break took place immedately adjacent to Gold Hill and some of the debris plunged into the cut, covering several tracks. This slide let down a small part of the jointed "mushroom-shaped" basalt of Gold Hill, which had rested to some extent on the strata that yielded. The fall from Gold Hill increased the load above the high bank and it "was followed on January 20, by a typical break by which the rock bluff which was holding back the upper mass of clay broke at or below the bottom level of the canal, completely filling the prism with clay and rock, reaching to 69 feet above sea level on the opposite or west side of the cut. The length of the prism so filled was 1,600 feet. Steam shovels were scarely able to keep pace with the movement, tracks were covered and disarranged **' As the movement continued, the clay broke farther and farther up the hillside." 20 (See pls. 7 and 8.)..
Following the break of January 20 the slide rapidly extended its area by additional breaks or slips of surface material until on February 12 it was estimated that 2,000,000 to 3,000,000 cubic yards of clay and rock were in motion. A good deal of this was surface soil and clay but the larger portion, probably, consisted of Cucaracha rocks.
20 Gen. Goethals' Supplement Canal Record, January, 1916, p. 6.
PANAM1A CANAL SLIDES-CROSS.
[MEMOIRS NATIONAL
[VOL.
By April the upper limit of the slide, near the southeast of Gold Hill, was in places 1,700 to 1,800 feet from the canal prism and had reached a water shed 500 to 600 feet above sea level. Much of this upper material consisted of a superficial red clay and soil such as had made up the earlier slides. Steps were soon taken to decrease the amount of this material which would otherwise be involved in the slide by sluicing part of it from the divide eastward, away from the canal slope. In this way about 1,000,000 cubic yards were removed from the head of the slide, during the year.
A fresh movement took place early in May some distance farther south of Gold Hill, and part of its material reached the main cut. As the rainy season progressed the slide material was rendered more and more fluid and flowed steadily into the cut. A special impulse was given to a part of it in June and the cut was filled to the 67-foot level on the west side. Another advance, which occurred in August, is represented in plate 9, as are also the general conditions adjacent to Gold Hill.
During the summer the shovels could not do more than keep pace with the flow so that onl October 10 the cut was completely obstructed for 900 feet, near Gold Hill and partially blocked for 1,100 feet farther south.
At this time the entire canal prism had been excavated nearly, or in places quite, to its proposed bottom, 40 feet above sea level. As the material of Cucaracha and Culebra slides was mainly soft it was decided that the most economical method of excavation in future would be by dredges. Accordingly the Gamboa Dike, which had held back the water of the Gatun Lake from Gaillard Cut, was blown up by dynamite on October 10 and the water, rushing through the gap thus made, reached the Cucaracha slide in two hours time. As the water reached the slide it was attempted to make a channel through it by dynamite explosion but the clay of the slide slumped back so quickly as to close the break. Later blasts gave similar results. A channel was soon made, however, and in the latter part of October removal of slide material by dredges was begun.6
The glacierlike form of the slide is exhibited in plate 8. The viscous character of the material which could make such a flow can be readily comprehended from the study of this view and the ineffectiveness of blasts in such material, vividly shown in plate 11, is not surprising.
For the rest of the. year 1913 the Cucaracha slide kept up its movement with occasional rapid advances, but on the whole the dredges gained on it and gradually cleared a channel along the entire front. The last stages in the removal of this great slide are graphically pictured in plates 53 to 56 of the annual report for 1913-1914.
West Culebra slide.-There are few references to special movements in the West Culebra area in 1913, either in the engineer's reports or in the Canal Record. In January a part of the slope below Zion Hill embracing 250,000 to 300,000 cubic yards, moved downward destroying railroad tracks on the upper terrace levels, but this material did not at this time reach the canal cut proper .2' A local sudden break in the bank of the cut in May caused some upheaval of the floor, but such disturbances were rare. (See pl. 17.)
During most of the year there was movement of some part of the slide into the cut,, but excavation progressed until the bottom level of the canal had been reached and for the most part the west bank had been cleared. With decrease in movement the slide material stiffened, so that the proj ecting toe of some portions could be removed without an immediate further advance of the slide.
The lightening of the west bank was carried on until December, at which time 8,797,990 cubic yards had been removed from the west bank, resulting in slopes of from 1 on 2.46 to 1 on 4.35. The slope was left with a series of benches from the edge of the cut to the steeper slopes of Zion and Culebra Hills, where the fissures had indicated the probable limit of slide movement. The result of the lightening work on both sides of the canal, as stated by Gen. Goetbals, "was that when the operations in the cut in the vicinity of Culebra were completed,
21 Canal Record, Vol. V, p. 165.
36
ACADEMY OF SCIENCES.] ITRCLSEC FLNSIE FGILR U. 3 XVIII.]]HITRCLSEC OFLNSIEOFGILR CU. 3
prior to the admission of water, the breaks in the banks and the upward movement of the hot-. tom had ceased entirely." 22
East Culebra slide.-The movements during 1913 in the East Culebra area were of impor-. tance, but there are few published records of particular movements. In January there was a break in the bank opposite Culebra 1,000 feet long, but it did not extend far back from the face of the cut. The Canal Record mentions that "on Thursday morning, June 12, another downward and upward movement occurred in the break in the east bank of the cut, which overturned a shovel, destroyed all tracks but two, and filled the drainage ditch." This movement left the line of cleavage in clear view in one section of the disturbed area, which now extends nearly back to the last level where terracing operations are in progress, and cracks have developed back of this line. The extreme limit of the break is now about 300 feet from the line of the canal." 23 This slide was near Gold Hill. Another smaller slide adjoining the basalt contact took place in December. A good view of this slide is given by plate 31.
On the whole the activity of the East Culebra slide, measured by material moving into the cut or by extension of sliding ground, decreased markedly in 1913. On the conclusion of lightening work on the east slide, in December, 6,533,924 cubic yards had been removed and the terraced slopes varied from 1 on 1.5 to I on 6.5.
While surface movements decreased there was an intimation of the deep-seated deformation of the rocks which resulted eventually in the great slide of 1915. According to Gen. Goethals, the fracture which now marks the rear of the East Culebra slide follows the line of a fissure which appeared in the latter part of April, 1913. This crack was in an old French dump, 1,300 feet from the canal prism. It was parallel to the canal and was not connected with cracks in the banks.,' There was no breaking up of the banks opposite nor any observed subsidence of the intermediate ground. The detection of the crack led to steam shovel work and sluicing to reduce the load where the crack had appeared. As no further opening of~ this crack appeared during the year it was concluded that it was not connected with a deep-seated movement.24
Hagan's side.-A very well defined slide on the east bank, about 1,200 feet north of the present East Culebra slide, began in February, 1913, the material involved belonging to the middle and lower part of the Culebra formation. It is now known as Hagan's slide but in the annual report of 1912-13 it was not distinguished from other slides on the east side of the cut. The first recorded motion took place on February 4, although the Canal Record states that warning cracks had been observed two months earlier. The slide started with a subsidence of 60 feet and a lateral movement of 80 feet into the cut. The break rapidly extended for 1,000 feet along the bank and 680 feet back from the east line of the canal. It was estimated that 2,000,000 cubic yards were set in motion. A very good representation of the zone of subsidence adjoining the fracture at the rear of the slide, on February 6, is given on plate 44, while the conditions in the cut on the same day, resulting from upheaval and lateral movement, are shown in plate 43. The general relations of the slide are brought out in plate 45.
On March 12-13 a further pronounced movement of this mass took place accompanied by an upheaval in the bottom of the cut for a distance of nearly 1,000 feet, with a maximum elevation of 30 feet. This over-turned shovels, destroyed tracks, and caused much other damage. A third impulse to this slide was given on April 16, again accompanied by an upheaval. A fourth development of the same character occurred on May 26 during a heavy rain (see p. 46), and still other similar movements took place in June. When water was let into the cut in October, this was one of the few points at which slide material projected into the canal (see pl. 47).
Other slides.-Except for a very small slip at Pedro Miguel, there was no new slide in the outer parts of Gaillard Cut in 1913. The following slides became inactive: Division office or Empire, north La Pita, Whitehouse yard, Buena Vista, and Haut Obispo. The Buena Vista is the only one of these which had renewed its activity up to the beginning of 1916.
22 Supplement Canal Record, January, 1916, p. 11. 24 Supplement Canal Record, 1916, p. 11.
23 Gen. Goethals, Supplement Canal Record January, 1916, p. 11.
80902
38 PANAMA CANAL SLIDES-CROSS. [MEORSNTOL
SLIDES OF 1914.
General statement.-The year 1914 witnessed a general decline in the slide activity except in the East Culebra area. No new slide areas developed and the Empire and Lirio slides became inactive.
The advance of material from the three great slide areas into the canal prism necessitated the greater part of all the excavation of the year in Gaillard Cut. For the fiscal year ending June 30, 84 per cent of the excavation was from slides and amounted to 2,635,902 cubic yards.
By midsummer the lessening movement of the Cucaracha slide led to confidence that the channel could be maintained along its foot, the only section where there appeared to be danger. The canal was therefore opened to commerce on August 15. Cucaracha slide answered to expectations and gave no further trouble, but a large new movement in the East Culebra slide area in October upset all calculations and further sliding closed the canal repeatedly, as described below. Warning of this movement had been given early in the year by the development of cracks at the rear of the slide but movement was so long delayed that a condition of stability appeared to have been reached.
The break of October 14 set in motion an amount of material far larger than the total excavation for the preceding year. It was of the greatest significance as again demonstrating the incompetence of the Cucaracha formation to sustain its own weight under the relations of slope existing in the Culebra district.
Cucaracha slide.-The Cucaracha slide mass which started in 1913 continued its glacier-like flow into the canal during the first half of 1914. "On June 30 the total area of the slide was 60.4 acres, 44.6 acres active and 15.8 acres without motion."251 The average daily advance at the face from the beginning of dredging operations to June 30, 1914, is given as 21 feet .26 The rate of flow was alternately accelerated and retarded and the dredges could not for some months do much more than remove the incoming material. Gradually the pressure on the slide decreased with the outflow of material, and the channel was widened and deepened until the passage of ships was deemed safe. When it became clear that the dredges were able to keep pace with this movement, there being no other obstacle, the canal was thrown open to commerce on August 15.
While the slide kept some dredges busy until October, it became inactive with the advent of the dry season in December. No new addition of note occurred during the year.
'West Oulebra slide.-There was no noteworthy movement of the West Culebra slide in progress at the beginning of 1914, and any small advances of slide material into the canal which may have taken place were too unimportant to be noted in the Canal Record. The superficial and deceptive appearance of stability for the slope below Culebra and Zion Hills, which existed when the work of lightening the load had ceased, was brought in question only by the development of certain cracks.
A noteworthy crack had existed for some years on the slope of Zion Hill. An inspection during the early, dry months of 1914 revealed the presence of several new cracks on the upper benches of Culebra Hill. "The cracks were suspicious, and to guard against possible contingencies it was considered wise to resume steam shovel operations. This was done and continued until July 1, 1914, when all trace of the cracks having been removed, excavation was stopped.
As the last steam shovel was being taken away from the west side the lowest bench gave way, the material moving into the cut; the amount was small and easily removed by dredges.'"21
A reference to the appearance of cracks in the West Culebra area, in the latter part of 1914, is included in the following statement by General Goethals: An examination of the west bank made subsequent to the break of October 14, 1914, on the east side developed the existence of cracks on both banks, and the question of resuming steax-shovel operations on both banks was seriously considered, but abandoned because the amount of work that could be accomplished by the shovels would be relatively small and the cost excessive. The breaking up of the banks
26 Ann. Kept. Isth. Canal Corn., 1913-14, p. 32. (Goethals.) 26 Loc. cit.
27 Gen. Goethals, Trans. Int. E ng. Cong. Vol 1, pp. 375-376, 1916.
ACDEY F cI.cs. HISTORICAL SKETCH OF LANDSLIDES OF GAILLARD CUT. 39
proceeded so rapidly that this conclusion has been fully justified. "21 The extent of this "breaking up taking place in 1914 is not indicated by any statements which have been noted.
East Culebra slide.-There was a slight renewal of activity in the East Culebra slide in January, 1914, but it soon subsided and until after the opening of the canal whatever advance of material into the channel took place was readily met by the dredges. Some signs of impending trouble were noted here, as well as on the west side of the cut, early in the dry season, in the appearance of cracks on certain levels back of the cut. Steam shovels were again set at work but they were removed in April, when the cracks seemed to have been "dug out."
On October 14-15, two months after the canal was opened to traffic, a section of the east bank extending 2,100 feet along the face of the cut and about 1,000 feet back from the center line of the canal settled down almost vertically, in some places as much as 20 feet. This block was bounded at the rear by an irregularly curved line at from 215 to 265 feet above the surface of the water in the canal, while to the eastward the relatively flat country was not much higher.
The surface of the upper portion of the sunken area remained for a time nearly parallel to its former position. Several benches preserved their relative positions though much broken up. On the canal front about 725,000 cubic yards of clay and rock moved into the canal prism within 12 hours, entirely closing the canal to navigation. Dredges attacked the obstruction at once and were able, by October 20, to restore a sufficient depth of channel for the passage of ships. The breaking up and sliding of this large mass into the canal continued at irregular rate, and a pronounced movement, with upheaval of the bottom, closed the canal from October 31 to November 4. For the remainder of the year the dredges were able to maintain a channel in spite of the steady advance of the slide.
Difficulty in maintaining the depth of the canal necessary for passage of ships was experienced and is referred to by Gen. Goethals as follows: "Surveys of the canal in the vicinity of the active slides were made daily and the channel dragged and marked prior to the passage of shipping. At times the channel shoaled so rapidly that it was necessary to drag after the passage of each ship, and in some instances such shoals required dredging before the following ships could proceed." 21Ilie does not indicate whether the shoaling is referrable to upheaval or lateral movement from the base of the slide.
SLIDES OF 1915.
Cucaracha slide.-During the dry season in the early months of 1915 the slide was almost at rest. The conditions at this time in the main portion of the Cucaracha slide are graphically exhibited in plate 12, and there has been but little change since that view was laken, on January 19, 1915. With the beginning of the rainy season the slide exhibited activity in certain parts. Before July 1 the whole slide became inactive. Only a little movement developed locally during the remainder of the year.
West Culebra slide.-After the great slide from the east side which blocked the canal in October, 1914, there was a gradual breaking up of a large area below Culebra and Zion Hills, but a movement en masse toward the cut was delayed for several months. (See pl. 18.) In the early part of 1915 the principal development was a progressive breaking up of the surface with cracks appearing at the rear farther and farther up the slopes of Culebra and Zion Hills, until one appeared at 480 feet above sea level, on Zion Hill. This is within a few feet of the present limit of the slide and near the summit. On Culebra Hill the upper cracks reached the 350 foot level and, as seen in the accompanying views of plates 19 and 21, approached near to buildings on the summit of the hill.
The crushing of underlying rock which was in progress during the first half of the year finally led to a pronounced movement into the cut on August 8, at the same time that a considerable mass broke away from the f ace of Zion Hill and settled down, producing the situation illustrated by plates 21, 22, and 23.
29 Ann. Rept. Isth. Canal Com., 1914-15, p. 27.
28 Ann. Rept. Isth. Canal Com-, 1914-15, p. 27.
40 PANAMA CANAL SLIDES-CROSS. [MfEMOIRS NATIONAL
The accelerated movement of the slides which closed the canal from September 4 to September 9 is not referred to in any way in the Canal Record of September 8, but in the number of Sepember 15 it is noted that "the greater part of the movement at this tim e is from the west side." "On the west side the break runs back 1,450 feet from the center line and the present edge on Zion Hill is at 585 feet above sea level." The east and west slides met on a line extending for 2,800 feet along the canal.
During the last three months of the, year the West Culebra slide was far less active than that on the opposite side, but at various times during each month a local advance for some section of the front was observed. Fortunately for dredging operations a considerable section of the southern portion of the slide, opposite the most active division of the East Culebra slide, became almost stationary and did not break up so quickly as other sections. This temporarily anchored part of the slide is well shown in plates 21 and 22, while plate 18 exhibits the brokenup condition of the face adjacent to the canal.
The gradual sliding en masse of the material in front of Zion Hill for several months after the sudden advance in August can be appreciated by an examination of plate 24, a view taken October 24. This view also shows the sharp southern boundary of the West Culebra slide.
East Culebra slide.-The movement of the large mass which started toward the canal in October, 1914, continued all through the succeeding year, with accessions of new slide masses at the rear and on the slopes of Gold Hill. The entire mass became more and more broken up into a succession of irregularly placed ridges, mounds, trenches, and sinks which continually shifted their relative positions from time to time, because of variability in rate of motion in different places. One of the slides adjacent to Gold Hill which occurred in January is represented in plate 33.
By constant work the dredges were able to keep pace with the advance into the canal, and thus maintain a channel for the passage of ships, until August, when renewed activity on the west side occurred and the combined slides completely filled the canal for a distance of several hundred feet.
The irregular movement of the Culebra slides from the time the traffic was blocked, August 7, to September 18, when the canal was officially closed for the remainder of the year, is shown by statements from the Canal Record. A channel was cleared on August 10, and the passage of vessels was possible until September 4, when the slides again closed in. The channel was reopened on September 10 for light-draft vessels, and by the 13th large ships drawing 30 feet or more of water made the passage. This condition lasted only a few days, and on Septmbr18 a break directly north of Gold Hill effectively choked the canal and this time the interruption continued until December 20, when six small vessels with maximum draft of 15 feet traversed the cut.
There was decided upheaval of the bed of the canal connected with the last slide, as illustrated by plate 34, taken September 21. The advance of the East Culebra slide was most rapid within a distance of 800 feet near to the base of Gold Hill. During October and November the work of all available dredges working at maximum capacity 24 hours a day was insufficient to restore a channel for any but very small boats.
The condition of the surface of the slide adjacent to Gold Hill at a time of rather rapid movement is well shown in plate 35, a view taken October 23. At this time the channel was almost completely closed by an advance at the base of Gold Hill, as seen in plates 36 and 37.
In November several rapid advances of the slide near Gold Hill completely filled the canal from bank to bank for a distance of 600 feet or more. The appearance at this portion of the cut after the dredges had cut nearly through this obstruction is illustrated by plate 38, taken on November 18.
To add to the embarrassment thus caused, the movement of soft Cucaracha beds adjoining the mass of Gold Hill removed the support for parts of the projecting basalt of the hill and large blocks of this massive rock fell upon the soft clay material. Part of this took place so near the canal that some fragments of basalt slid into the water, while others were perched at various
ACADEMMY OF SCIENCES.]HITRCLSEC OFLNSIE OFG LADCU.4 xv"'i.]HITRCLSEC OFLNSIE OFGILR CU.4
elevations not far from it. This fall of basalt is similar to that which occurred on the border of the Cucaracha slide in 1914.
During the month of December the rate of advance of both Culebra slides decreased materially and the dredges were able to open a channel for small boats which was not again closed, although the depth of the channel was subject to many fluctuations. In plate 39 is represented the appearance of the cut on December 23.
SLIDES OF 1916.
General statement.-The Culebra slides continued to move toward the canal during the entire year, fluctuating in their advance from time to time, but with a gradual decrease in the rate, and even an apparent suspension at times. The dredges removed the material as it reached the canal prism. The full dredging force was at work in three 8-hour shifts every day until December, when dredging was suspended on Sundays and holidays. The amount of Culebra slide material removed from January 1 to December 1 was 9,521,781 cubic yards, leaving, an amount of moving material estimated at 6,347,656 cubic yards, which the engineers think it will be necessary eventually to remove.30 The excavation for the year was thus nearly twice as great as the estimated remainder of the slides. Accepting this estimate as correct, it is plain that without substantial additions to the slides the dredges can easily remove the remainder of the slides within the year 1917, if it comes within reach.
On April 15 the canal was reopened to commerce and it was not again obstructed except for a few days in August by a talus slide in the Cucaracha area. For some months the channel was subject to frequent shoaling by upward movement of the bottom so that it was necessary to drag the channel and perhaps do some dredging before the passage of each ship. Gradhially the bottom of the canal became more stable and during the last three months of the year there was little appreciable shoaling and a broad channel was maintained without difficulty.
In order to secure a navigable channel without interruption to dredging, it was decided to widen the prism of the canal to 500 feet, by an addition of 100 feet on each side for almost the entire lenth of the Culebra slides. Nearly all of this increased width has been secured. Plate 50 represents the progress made in widening the canal on July 14, and the general condition of both Culebra slide fronts.
Outside the area of the Culebra slides there was no important slide movement. The renewed movements in the Cucaracha and Buena Vista slide areas were of small volume and were readily handled as described below. No new slides developed.
.*est Culebra slide.-The West Culebra slide area changed greatly in appearance during the year. While much of the front portion slid into the canal and was removed, a very notable feature of the movement was the nearly vertical sinking of the rear part of the slide, adjoining the scarps of Culebra and Zion hills. This subsidence has been apparently much more pronounced than has the corresponding movement at the rear of the East Culebra slide. According to the personal statement of Col. Harding, made to the writer, the surface of the West Culebra slide has assumed an average grade considerably less than that of the East Culebra slide. This marked and gradual subsidence can perhaps be explained on the assumption that the greater part of the upheaval or shoaling in the canal was connected with the subsidence on the west side.
The surface of the West Culebra slide, which in the latter part of 1916 was characterized by several areas resembling fault blocks, with slightly tilted and almost unbroken surfaces, as illustrated in plates 21, 22, 23, and 24, has become very irregular. Plates 25 and 26 exhibit this rough surface and also the steep scarp at the rear of the slide as they were in July, 1916. Both features have become greatly accentuated since those views were taken.
30 Ann. nept. Gov. of Canal Zone, 1915-16, p. 302. The estimate given by W. G. Comber, superintendent of dredging division, reduced by the yardage removed after June 30.
42 PANAMA CANAL SLIDES-CROSS. [MmOIRtS NATIONAL
The sharp pinnacle of basalt which broke away from Zion Hill and is illustrated in plate 21 has disappeared, having been broken up by blasting as it slid down the slope and much of its material has entered the canal and been removed.
A slight enlargement of the West Culebra slide was due to sloughing off of small masses at the rear and on the south side.
East Culebra slide.-The principal feature of the East Culebra slide movement was a general advance at irregular rate toward the canal, where the dredges have gradually gained on the slide. There were some marked subsidences of the rear portion and the area of the slide has been increased to a small extent by the breaking down of the rear banks. No large additions have occurred at any one time.
"In May and June there were two general movements of the east bank within the sliding area for a length of about 2,500 feet. These movements were accompanied by a vertical settlement of the bank arnd a narrowing of the channel opposite. In some places the vertical settlement was as muc h as 25 or 30 feet." 3' The shoaling of the channel accompanying these movements was confined to its eastern half and the material was soon removed.
The body of massive rock called "Gibraltar" by the workmen, which became detached from the north face of Gold Hill near the canal, caused the engineers much anxiety for a time, since it was so situated that a sudden acceleration of its movement might bring it into the channel from which it could be removed only after its hard rock had been broken up by blasting. Fortunately this body has moved slowly, at intervals. Its mass has been reduced greatly by blasting and as it approached the canal the dredges attacked its northern base apparently diverting its movement somewhat to the north. On December 20 the Canal Record reports that Gibraltar "extends into the channel about 110 feet from the prism line, for a distance of about 200 feet along the axis of the canal." It then rose only about 30 feet above the surface of the water and its complete removal within a short time is certain.
Cucaracha slide.-The Cucaracha slide remained inactive during the year but the slipping into the canal of some debris from Gold Hill, which was within the limits, of the former Cucaracha slide, gave rise to the report that there had been reftewal of movement in the great slide. Gen. Goethals' monthly report for August, 1916, is inadvertantly misleading in referring to this movement, when he says: "Cuicarachia slide, which had been quiescent for the past year, began to move again on August 2, and was very active from August 24 to 3 1." 1 From various references to this movement in the Canal Record and personal advices from Col. Harding, to the writer, it is certain that the material which slid into the canal in August was the pile of talus from Gold Hill, represented in plate 12, augmented by further debris from the cliffs.
This debris doubtless rested on a surface of soft Cucaracha rocks inclined somewhat toward the canal. The official map of June 15, 1916,~ shows that this debris pile then formed a mound rising 45 feet above the canal with abrupt slope toward the water. The weight of this material pressed the lower blocks of basalt or breccia into the canal. By August 31 the mass had advanced slightly more than half way across the channel and passage except for small boats was interrupted until September .
The removal of this talus slide was difficult because some of the blocks beneath the water were so large as to require drilling and blasting to reduce them to fragments that could be handled by dredges. After removing this slide material the dredges widened the canal to 400 feet for a distance of several hundred feet south of Gold Hill, providing a space where talus from the hill can accumulate without interfering with the ship channel.
IAccording to Col. Harding there has been no break in the cliff of Cucaracha beds south of Gold Hill, seen in plate 12, and there is no known reason to assume that the advance of this talus pile into the canal was in any way connected with a renewed movement of the Cucaracha slide proper.
31 Letter of Col. Harding to Chairman Van ise, of date Oct. 30, 1917.
32 Canal Record, Vol. X, p. 58, 1916.
33 Ann. Rept. Gov of Panama Canal, fiscal year 1915-16, plate 96.
ACADEMY OF SCIENCES.] ITRCLSEC FLNSIE FGILR U.4 XVIII.]HITRCLSEC OFLNSIE OFGILR CU.4
Buena Vista slide.-This slide, which is near the northern end of Gaillard. Cut, on the west side, renewed its activity for a short time, but was quickly stopped, according to the statements of Gen. Goethals. The old slide at Buena Vista showed signs of new life in December, 1915, and' on January 1, 1916, a crack had developed, starting at station 1559 and running over the top of a small hill 300 feet west of the prism line at station 1563 and joining the canal again at station 1567. The material was stiff clay and soft rock of the Las Cascadas formation. On January 3, 1916, the hydraulic grader began operations, starting at a point about 50 feet back of the crack and cutting a uniform slope from the point of beginning to the base of the slide at the water's edge. This method stopped the movement and the material has remained quiescent during the rest of the fiscal year." y 3
REVIEW.
The general course of slide activity in Gaillard Cut has borne a marked relation to the progress of the canal* excavation. As soon as the sides of the cut reached a considerable height slides began at certain weak places. During the years 1908-1912, while the bottom was being lowered the slides became most numerous. The full depth of the canal was reached in 1912 over a large portion of the cut and since that year but three new slides have begun.
The decline in the number of separate slide movements lagged somewhat behind the excavation, of course, but within two years after the desired depth of the canal had been reached the slides outside the Culebra district had stopped practically. These relations may be grasped most readily by the aid of the accompanying table, the data for which were taken from the statistical Appendix C.
Table showing development of slides during the period of canal excavation.
Slds Total
Y~ear. Slides Slides number of
began. bcme active
inactive, slides.
1907...............4 0 4
1908------------- 5 0 9
1909...............3 1 12
1910...............4 0 15
191..............1 2 16
1912............. 6 5 19
1913............... 7 16
1914............. 1 5 10
1915...............0 2 5
1916...............0 1 3
The units of this table are distinct named slides. Some are large and some are small. The general significance of the relation brought out will be clear to everyone. The banks have reached a condition approaching stability except in the Culebra district. New slides beyond that area, which may appear in the future, will be due to slow development of some weakness under conditions which have existed for at least two or three years.
34Ann. Rept. Gov. of Pan, Canal, fiscal year 1915-16, p. 27.
9
Appendix B.
THE GEOLOGY OF THE PANAMA CANAL WITH SPECIAL REFERENCE
TO THE SLIDES.
By DONALD F. MACDONALD.
Introductory.. ----------------------------------------------------------------------------------46ge
Surface configuration controlled by rock formations--------------------------------------- --------------- 46
Rock'formations within the Canal Zone-------------------- -------------------------------------------- 46
Brief descriptions of the formations---------------------------------------------------------------- 47
Fragmental volcanics, roughly bedded --------------------------------------------------------- 47.
Bas Obispo formation-------------------------------------------------------------------- 47
Las Cascadas agglomerate ---------------------------------------------------------------- 47
Bedded rocks ------------------------------------------------------------------------------- 47
Bohio conglomerate ---------------------------------------- ------------------------------ 47
Culebra formation ---------------------------- ---- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 47
Cucarach; formation ------------------- --------------------------------------- ---------- 49
Emperador limestone -------------------------------------------------------------------- 49
Caimito formation ----------------------------------------------------- 49
Gatun formation ------------------------------------------------------------------------- 50
Panama for nation ------------- ------------------ -------------------------------------- 50
Caribbean limestone ------------------------00------------------------- 50
Pleistocene formations-------------------------------------------------------------------- 50
Igneous rocks------------------------------------------------------------------------------- 50
Basalt ----------------------------------------------------------------- ---------------- 50
Andesite------------------------------------------------------------------------------- 50
Meta-tuffs ------------------------------------------------------------------------------ 51
Volcanic mud flows ------------------------------------------------------------------ --- 51
Other igoneous rocks---------------------------------------------------------------------- 51
Structural geology in relation to the slides------------------------ ------------------------------------- 51
Sections along Gaillard Cut showing geology in slide area--------------------------------------------- 52
Figures 2-5. 4
(MEmNIQRS NATIONAL
[ VOL.
THE GEOLOGY OF THE PANAMA CANAL WITH SPECIAL REFERENCE TO THE SLIDES.
By DONALD F. MAcDONAL D.
INTRODUCTORY.
The rocks traversed by the Panama Canal consist of Tertiary sediments and Tertiary volcanics. The latter occur mostly as basic necks, dikes, and masses, chiefly in the central section, but sporadically on the Pacific slope; the former, in the main, constitute the terrain between the seacoasts and the central hills, and are especially predominant on the Atlantic side.
SURFACE CONFIGURATION CONTROLLED BY ROCK FORMATIONS.
The surface configuration of the Canal Zone is controlled by the rock formations. As there are two chief types of rock-the soft friable sedimentary and the hard volcanic; so there are two chief types of topography-the hard volcanics, occurring as they do in small masses cutting softer rocks, form innumerable angular peaks sprawled over the central or watershed zone, and to a lesser extent on the Pacific slope, and the soft and easily eroded sedimentaries form the lowlands and the valleys.
The surface configuration in the hilly areas is most striking in its irregularity. The original angular forms of the intrusions are yet partially preserved in the forms of the hills which have been shaped from them by erosion. None of the peaks are over 1,000 feet in elevation and, standing in every direction as they do, they present some likeness in configuration to enormous crosswaves at sea. The similarity is heightened where the green jungle mantles the slopes. The lowland areas, especially near the coast and along the lower valleys of the rivers, contain many black mud swamps.
That part of the canal which traverses the divide or watershed zone of the Isthmian land mass is called Gaillard Cut. It is about 81 miles long and is by far the deepest and heaviest cut along the whole length of the canal. It has a minimum width in the bottom of 300 feet and an average depth of over 100 feet, a maximum depth of over 400 feet, and an average depth in the Culebra section, where the biggest slides occurred, of 200 feet or more. Where the slopes of the cut attain maximum height (where the cut passes through the flanks of the highest hills which the canal encounters) there has, fortunately, been no serious slide. This is due to the fact, already stated, that. the hills are in the main composed of hard igneous and metamorphosed rocks, which buttress back the softer rock and tend to prevent sliding. However, down on the flanks of these hills, where the cut has a depth of 200 to 300 or more feet and where it encounters soft clayey rock, maximum sliding developed.
ROCK FORMATIONS WITHIN THE CANAL ZONE.
Within the Canal Zone nine different sedimentary formations were recognized, together with two formations of fragmental volcanic rocks and agglomerates, and six or more different types of massive volcanic rock.
Correlation of the different beds across the Isthmus was very difficult, owing to the soil cover, the jungle mantle and the general scarcity of outcrops. Especially was this found to be true of the extensive flat area now occupied by Gatun Lake. The correlation here presented is, therefore, tentative and may be changed somewhat when studies, now under way, on the fossil collections made, are completed. Figure 2 shows the present correlation -from the Atlantic to the Pacific, and figure 3 shows the geological column or rock succession.
46
Cartbbeazi Sea
0
co-.Toro Point
Monkey Hills ~2 '~ Tier HUiLl
CAlM/TO
Rio Chaqre6
'ZA- OASPODA
Gold 1Hfil (~
: K"PA7L~I1So
P4Ii .3AO MIOUZ7.
3 NA> Ancon HIl
(l~Of tzriania) *
a
ACDEY FSENE.] GEOLOGY OF THE PANAMA CANAL. 47
BRIEF DESCRIPTIONS OF THE FORMATIONS.
FRAGMENTAL VOLCANTOS, ROUGHLY BEDDED.
Bas Obispo formnation.-The Bas Obispo, the oldest formation, consists of fragments of andesite blown out of old volcanic vents and later consolidated and cemented into solid rock, massive tuffs and breccia, by silicious and limy cementing material. Pyroxene-andesite and some hornblende-andesite are the principal rock varieties of this formation. Its base was not found, so its thickness is unknown. It is separated from adjacent formations by fault contacts. It outcrops extensively at Bas Obispo near the north end of Gaillard Cut, and near Old Panama, and small outcrops rise above the alluvium near Miraflores and Diablo Ridge.
About 7,000 feet (2,134 meters) of the north end of Gaillard Cut has been excavated in this formation. It is relatively hard and tenacious, and except locally where sheared by faulting, has stood fairly well at steep angles, even where the slopes of the cut are high. From these faulted places masses of loose rock have fallen, but not enough to be classed as important slides.
Las Cascadas agglomerate.-The Las Cascadas agglomerate also consists of volcanic dbris partly consolidated, but much less so than the Bas Obispo formation. It contains large and small subangular fragments in a fine-grained matrix of volcanic clay and tuff, and some solidified ash or lava-mud flows, as well as flows and dikes of grayish andesite, and considerable andesitic breccia locally interbedded with the agglomerate.
This formation outcrops along the canal between Empire and Las Cascadas over a distance of 2 miles. It is of very variable constitution and is much softer and more friable than the Bas Obispo formation. Several hundred feet in thickness of it are exposed, but its base is not known.
Some of the beds and lenses of this formation are relatively fine grained and impervious, and have locally a soapy feel. The contacts between the beds dip toward the west at angles of from 200 to 350; thus the rocks on the east side of the canal dip toward the cut. Because of this canalward-dipping structure, and the impervious, weak and slippery character of some of the beds, 10 slides, having a combined volume of over 31 million cubic yards, developed on the east side of the cut. The rocks on the west side are just as weak, but owing to the fact that they dip away from the canal, only 5 slides, having a combined volume of two-thirds of a million cubic yards, developed on that side. The slides in this formation, and their yardage, are given in the table, page 64, Appendix C.
BEDDED ROCXS.
Bohio conglorerate.-The Bohio conglomerate does not outcrop along the Gaillard Cut. It consists of andesitic and dioritic bowlders, cobbles, pebbles, and sandy material, arranged in beds which are separated from each other by beds of sandstone and argillite. Several hundred feet of this formation are exposed, but its full thickness is unknown. It outcrops extensively in the vicinity of Bohio and near Caimito Junction.
Culebra formation.-The Culebra, a shallow-water marine formation, contains an upper and a lower member. The lower (a) consists of dark, well-laminated beds of soft shales, marls, and carbonaceous clays, with some pebbly, sandy, and tuffaceous layers and a few thin beds of lignitic shale which locally show fossil plant remains. The upper member (b) consists of beds and lenses varying from sandy limestone to calcareous sandstone, a few inches to 10 feet thick, separated by partings of dark carbonaceous clays and fine bedded tuffs. These limy beds are locally rich in foraminifera.
The dark, carbonaceous, friable, shale beds contain a few minute angular grains of quartz and feldspar, the latter mostly plagioclase, somewhat sericitized, and a few cloudy areas, the remains of ferromagnesian minerals, all in an exceedingly fine-grained, dark-colored matrix. Some of the coarser grained beds are made up mostly of small grains of andesitic material, considerably altered. The limy beds contain a few minute angular grains of quartz and feldspar and are rich inorganic remains, largely foraminiferal.
A faulted svnline near Las Cascadas shows this formation to rest on the Las Cascadas agglomerates. its thickness, however, is unknown, but is thought to be 500 or possibly 600
[MEumoiS NATIONAL
[VOL.
GEOLOGIC COLUMN
CANAL ZONE FORMAIQNS. .PLEIS /i-cre S wsan ,t TOCENE Mujjjjjlljan[d Sizesor
ii 0 Crbbeari f $/w/, rna'r/ cz'
('10- Qcrr ?JLiqiui colored ztalf CEt4E lfij 'or77lior? by'ds, arji//de etc.
m18-b be
t'1I0- j ..C Siidndtc77e
0CEWE b Liry c~qy/c ere7/
~ ~/ 6r eLofe 7 lycoral/s
...CLcardi'C/?& )/M/25sive so?
~ Ca /eZ?-,? fclrne cdrbhal~'e,
~ ~ 3 EO/?i08a'd5tone &td
L a Z~s Cuscrdds J Ii'/cdnzc caq/oerze
a-Py10 -0 e ze tuff.s'and 9va F/ca-s
Ba, O 1*5, 0VO/cd7 I' t2Fr Forz?7c77677 rzd b2(ecc1'ia
6 3 -i/t I17led
I
Me/.-c-F b z2& V cirite
FI. 3.
48
OF CIECIe.] GEOLOGY OF THE PANAMA CANAL. 49
feet. Locally, it gives off a little natural gas, and in some restricted areas it shows slightly bituminous shales.
The Culebra formation outcrops in Gaillard Cut at and north of Culebra and near Paraiso and Pedro Miguel. In spite of its -unstable character, only three large slides, the Lirio, New Culebra, and Hagan slides (table, p. 64, Appendix C) developed in it. It was less given to sliding than other equally friable formations because (1) it is well bedded and relatively horizontal, hence there is practically no tendency to slide along bedding planes; (2) it is relatively porous and does not retain water as the more clayey formations do; (3) in general these rocks develop less placticity with movement than seems to be true of the Cucaracha rocks (in which maximum sliding occurred); (4) the thin beds of sandy limestone and limy sandstone, in the upper part of it, are fairly resistant and where they outcrop in the slopes are strengthening factors against crushing and deforming movements. This upper and stronger member of the formation is about 150 feet thick. The slides which occurred in the f ormation had their basal or lower zone of movement, in general, below this member, or where these beds were cut and much weakened by faults. The middle and lower parts of this formation (350 to 450 feet thick) are relatively weak and friable.
Certain local areas of the Culebra formation became heated on exposure to the atmosphere through drilling or blasting. This I heating was due to the oxidation of finely divided pyrite. It was thus necessary to test the drill holes in certain areas to determine whether they had become hot and therefore dangerous to load for blasting.
6Cucaracha formation.-ln the Cucaracha formation most of the big slides developed. The formation consists of a greenish gray to grayish green, slightly indurated, very fine-grained, clayey rock which develops placticity on being kneaded. Throughout this massive greenish rock'are some beds and lenses, generally not more than a few feet thick, of red oxidized rock, apparently representing old weathered surfaces. Three different lignitic shale beds, up to 4 feet thick, are present, as well as a light-colored andesite (almost dacite) flow, 10 to 20 feet thick and exposed for 1-1 miles. The basal bed or stratum of the formation is loosely cemented gravel 10 to 20 feet thick, consisting of chert, meta-shale, and basaltic pebbles. Lenses of similar gravel with some sandy tuffaceous beds were noted in its upper part.
The origin of these rocks seems to be that they were deposited on land with locally swampy or shallow-water conditions which gave rise to the lignitic shale beds. The material of which the formation is composed seems, in the main, to be very fine grained, somewhat basic, volcanic ash, considerably altered, perhaps hydrothermally, but only locally weathered. It is probably 500 to 600 feet thick.
Though, in the main, very fine grained, these rocks have a high content of water (see p. 62, Appendix C) and of sof t smooth finely f oliate minerals (see p. 58, Appendix C) which have a soapy feel. Owing to its natural weakness, its high water content, the fact that it is cut by many faults, and that it chiefly forms the slopes in the deepest part of the cut, most of the large slides developed in rocks of this formation.
Eimperador lirnestone.-The Emperador limestone is a light-colored to light-yellowish, fairly pure limestone, about 50 feet or more thick, which contains many fossils, especially corals. It outcrops northwest of Empire, south of Las Cascadas, on the relocated line of the Panama Railroad near San Pablo, near Frijoles, and in the swamp southeast of Diablo Ridge. A wide area of linestone near Alhajuela may be of the same age or younger. It has very little relation to the slides, but where it forms a small part of the canal slopDes near Las Cascadas it is a strengthening factor against slides.
Caimito formation.-The Caimito formation does not outcrop in Gaillard Cut and so has no relation to the slide problem. It consists of upper and lower sandstone members with a calcareous agglomerate member in between. The sandstone members are soft, argillaceous, and grayish to yellow in color, and locally weather into somewhat spherical fragments. The
I MacDonald, Donald F., Heating of local areas of grou nd in Culbra cut, Canal Zone: Science, May 3, 1912, pp. 701. (N. S. Vol. xxxv, No. 905.)
50 PANAMA CANAL SLIDES-MACDONALD. [ VOIIR ATNL
calcareous member contains fragments of much decayed basic material which locally gives a bright green stain to small patches of the rock. The formation outcrops at Bald Hill north of Miraflores, extensively at San Pablo and on the relocated line of the Panama Railroad near New Culebra. The thickness of this formation is unknown, but must be several hundred feet.
Gatun forrnation.-The Gatun formation does not outcrop in Gaillard Cut, so has no relation to the slides. It consists, of three members: The lower, containing fossiliferous argillites, soft sandstones, and some conglomerate; the middle member, mostly fine, soft sandstone, containing a few fossils; and the upper member, made up of light to creamy-gray colored argillites. The fineness of grain of this rock renders it relatively impervious to ground water. This fact and its sandy character make it an excellent foundational site for the Gatun Dam and Locks. Its thickness is probably over 1,000 feet.
Panama formation.-The Panama formation does not outcrop in Gaillard Cut. It is a light-colored, well-bedded tuff with some argillitic material. It outcrops extensively from Miraflores to Panama and is relatively porous, somewhat jointed, and slightly friable. Its exposed thickness is several hundred feet.
Caribbean limestone.-The Caribbean limestone does not occur near Gaillard Cut. It is a soft, sandy, fragmental limestone, locally a coquina or shell marl, It outcrops at Toro Point, west of Gatun Dam, and at the mouth of the Chagres River. In most of these outcrops it forms low bluffs. It is the rock from which Fort San Lorenzo was built and was quarried from a convenient outcrop on Toro Point for use as a hearting for the breakwater there. Thickness, 100 feet plus.
Pleistocene formations. -The Pleistocene formations have practically no relation to the slides. They consist of: (a) Swamp formations, black soil and silt, filling old channels to depths of 375 feet (114.3 meters) below present sea level; (b) river gravels up to 10 feet (3.05 meters) above present normal river levels, and old sea beaches 6 to 10 feet (1.83 to 3.05 meters) above present beach level; (c) bars, beaches, and present river alluvium.
IGNEOUS ROCKS.
Within the Canal Zone there are six or more types of igneous rocks. Some of these have no relation at all to the slides, so will be given very brief mention. Four types, the basalts, the andesites in part, the meta-tuffs, and the volcanic mud flows were encountered in canal excavation and have some relation to the slides. They will, therefore, be briefly treated under separate headings in the order of their importance.
Basalt.-Basalt, in the main of the nonolivine-bearing variety, dark, fairly fresh, and medium grained, forms many cores, flows, and dikes along Gaillard Cut. Some of the dikes show diabasic structure, and locally olivine is present. Basalt outcrops near Empire, Culebra, Pedro Miguel, the top part of Gold Hill, the hills near Paraiso and Rio Grande; it also forms a hill near the Panama Railroad 2 miles north of Montelirio. In fact, most of the steep hills and ridges within the Canal Zone, except Aucon Hill, are formed of this rock. It is notable that most of the basalt intrusive masses are relatively small and very much jointed. Locally, dikes and small intrusions of basalt have, to some extent, acted as buttresses to hold back portions of the sliding material. However, owing to weakness from excessive jointing they were not as efficient in this respect as their size and position might lead one to expect. This was especially true of the basaltic intrusions which for a time held back the Cucaracha slide, but which finally collapsed under excessive stress from the sliding material.
Andesite.-Andesite, fine-grained, light gray to dark in color, mostly of the augite variety and fairly fresh, is known at several places in the Canal Zone. Necks and dikes of andesitic rock cut the Las Cascadas agglomerate and locally act as strengthening factors in the slopes of the canal. An extensive andesitic flow, which is almost dacite in character, interbedded in the Cucaracha formation, was not much of a strengthening factor in that formation because of its thinness, 10 to 20 feet, and its jointing. Other andesitic masses are numerous, but have no relation to the slides.
ACDEYIII INCD.] GEOLOGY OF THE PANAMA CANAL. 5
Meta-tuffs.-Under this heading, are classed the metamorphosed tuffs, agglomerates, and breccia-tuff masses that form a part of Gold Hill, Contractors Hill, Office Hill (Culebra), the tuff and breccia at Paraiso, Empire, and other places. Some of this meta-tuff may in reality be a part of the Obispo formation, faulted relatively upward to its present position. Associated with practically all of these breccia-tuff masses are basalt dikes. Both the Gold Hill and the Contractors Hill masses have been faulted downward probably more than a hundred feet since they were intruded by basalt dikes. In all cases where the contacts of the consolidated tuff masses were exposed they were found to be faulted. The evidence indicates that these masses may have been pushed upward as somewhat metamorphosed and toughened caps on top of basalt plugs or cores. They have acted as strengthening pillars to buttress up some of the sliding areas in Culebra Cut.
Volcanic mud ftows.--Jn addition to the igneous rocks above enumerated, there are several consolidated lava-mud flows in the Las Cascadas agglomerate. These locally show columnar structure. On exposure to the atmosphere for a few years they weather and crumble very considerably.
Other igneous rocks.-The igneous rocks found within the Canal Zone, but which have no relation to the slides, are granodiorites, found among the Chagres River gravel, thus indicating outcrops of this rock in the Chagres River drainage basin. This rock also forms Cocovi Island, a very small island in Panama Bay. Diorite in small masses occurs at Point Farfan, opposite Balboa, and among the gravels of the Chagres River. Rhyolite, of almost dacitic composition, forms Culebra and Naos Islands and Ancon Hill. This rock is fairly hard and tough but much fissured and jointed. It was used for the concrete work of the Pacific locks.
STRUCTURAL GEOLOGY IN RELATION TO THE SLIDES.
The structure of Panama, broadly considered, is anticlinal, with the axial zone of the anticlinorium punctured and pushed up by igneous intrusions. Near the intrusions the dips may be fairly steep locally, while out toward the low-lying coasts they are comparatively fat except for local folds.
A downf old, or syncline, of the bedded rocks trends across Culebra Cut near Gold Hill. (See section, fig. 4.) This brings the stronger limy sandstone beds well below the bottom of the cut for a horizontal distance of a mile or more and leaves only the material with which this syncline is filled, the weak argillaceous rocks of the Cucaracha formation, to form the slopes of the cut where it is very deep. This region is the locus of the greatest slides.
The many vertical movements of the Isthmian land mass and differential stresses have caused numerous faults in the rocks (see sections, figs. 4 and 5), and where these shear zones trend across the canal, they are weakening factors in the slopes and help to promote slides.
Jointing and fissuring cause most of the smaller masses of igneous rock to break out in very small pieces when blasted. In getting crushed rock from Ancon quarry, this feature saved the United States hundreds of thousands of dollars in blasting and crushing costs. In other cases, however, this excessive jointing which characterizes the basic intrusions of the Canal Zone, probably owing largely 'to their comparative smallness, militated against canal construction. A case of this kind is that of the dikes and flows of hard basalt which, for more than a year in 1912, held back the Cucaracha slide. They seemed large and heavy enough to hold it back permanently, but when the cut was brought to final depth in front of them, the accumulated stresses of the loose material behind the dikes caused the latter to be sheared off where they had been greatly weakened by joints and cracks, and thus the slide took on renewed action.
Such intruded masses of relatively tough rock, as Gold Hill and Contractors Hill, act as piers and buttresses to strengthen the weak slopes against sliding. However, the overhanging brows of these hills and of Zion Hill slipped off in some cases, owing to excessive jointinig, where the softer supporting rocks were removed from them.
52 PANAMA CANAL SLIDES-M NACDONALD. [ VEORSNTONL
SECTIONS SHOWING THE EAST AND WEST SIDES OF GAILLARD CUT BEFORE
MARKED SLIDING OCCURRED, FROM STATION 1720 AT EMPIRE TO STATION
1855, A THIRD OF A MILE NORTH OF PARAISO.
D. Basic dikes and intrusions; mostly basalt.
D-1. Peripheral dike which came up around the edge of the Gold Hill -mass and which, in the horizontal section, has
a curvature of 90 and a thickness of from 35 to 100 feet. D-2. Cap of basalt which forms the top of Gold Hill. D-3. Basaltic dikes which, for nearly a year in 1912, held back Cucaracha slide, but which finally, from excessive
pressure and through weakness from much jointing, sheared off, thus giving renewed activity to the slide. D-4. Small basaltic and diabasic dikes which cut the Cucaracha formation. D-5. Basaltic mass which outcrops on the slope some distance back from the cut and which, though much jointed, is a
considerable streng-thening, factor in the slope. Soft clayey rocks from farther back on the slope have locally
slid down over this mass.
D-6.. Basic ridge, mostly basalt, which curves irregularly around the head of Cucaracha slide, and which nearly
joins with the Gold Hill mass, thus forming an irregular hard rock rim which has limited the area of the slide. The rim is not quite continuous, but contains a few soft rock gaps where the numerous basic intrusive
masses which form it failed to make contact with each other.
D-7. Intrusive mass of basalt which appears to have laccolithic form. It has tilted the beds, but has achieved relatively little metamorphism of them, though some baking for a distance of a few feet was accomplished. D-8. Basaltic and diabasic dikes intruded into the meta-tuffs. C. Cucaracha formation-500 to 600 feet thick.
C-i. Loosely cemented gravel conglomerate which forms the base of the Cucaracha. C-2. Massive light greenish argillaceous rock of the Cucaracha formation, soft, friable and fine-grained; contains some
beds and lenses of the same type of rock in which the iron has been oxidized to a reddish color; also a few sandy
and tuffaceous beds.
C-3. Lignitic shale beds in the Cucaracha formation, from 1 to 4 feet in thickness. C-4. Andesitic lava flow now interbedded in the Cucaracha formation. This flow is 10 to 15 feet thick, light gray in
color, locally somewhat brecciated, and shows some little inclusions of dark material which seems to be charred
wood.
C-5. Fine conglomerate bed in the Cucaracha formation 1 to 3 feet thick. B. Culebra formation-500 to 600 feet thick.
B-i. Lower part of Culebra beds; dark, sandy, well-laminated shale, very soft, and contains considerable organic
matter.
B-2. Light-colored tuff bed locally overlapping Culebra bed. B-3. Local overlap in lower Culebra beds.
B-A. Fine conglomerate or. cemented gravel bed in the lower Culebra beds. Contains many fossil oysters; 3 to 5 feet
thick.
B-5. Much like ''B-i," but seemingly less sandy, better laminated, and with perhaps more carbonaceous material.
Fossiliferous; fossil plants.
B-6. Beds and lenses of liiny-sandstone and sandy-limestone in the upper part of the Culebra formation. These are
from a few inches to several feet in thickness and are separated by partings of well-lamninated dark friable
shale from a few inches to 10 feet thick. Fossiliferous; many foramiinifera; 150 feet thick. A. Meta-tuffs and breccias.
A-i. Meta-tuff near Empire. Well consolidated, dark, somewhat Inassive andesitic tuff; cut by basaltic and diabasic
dikes.
A-2. Andesitic-breccia mass or dike.
A-3. Meta-tuff, which forms the core mass of Gold Hill. Shows good bedding and, locally, is well laminated. Very
much tilted and folded and near the outer margin has been broken into coarse blocks and subsequently recemented into mosaics of coarse breccia.
A-4. Meta-tuff which forms Contractors Hill. Shows bedding in the upper part, but is fairly massive and contains
some angular diabasic fragments in the lower part. Bedding less distinct than most of that of the Gold Hill
meta-tuff and less tilted and folded.
F. Faults; where the faults extend across the canal they are given the same numbers on each section.
600 NORTH GOD R 550-D 506- Empire+ 450103-++ + y+ + + C ~ 4
300 A,F, A-,B, BZ P 4 ~F4 B5B601 F5 C 0C0? C3 3 0 zsdr
200 .
E'-A LEVEL. 70 7O701750 1760 170FI79 03180 4
FIG4 Section showing the east side of Gaillard Cut before marked sliding occurre
500 NORTH
400] 04F13 306- D8A, Az B, F, BF2 B 5 F3 F4 Bs Ps FqCI z C3020 0FCZ 9 0 0 zsd
s00
FrG. 5. Section showing the west side of Gaillard Cut before marked sliding occurre
00
08 OUTH -900, + 4+1 + ++4C 50' 4s3 2 Cs CC3 Ci,Ps Bs5 D7
3OUTH 200,
445
440
t 35 0 ,
3400
2-50
4 z+F 16 04 0 13 0 7Fis C05 C50,1 -.200
850840SEA LEVEL 0 o
d. 80698-4-4. (Fa-ce' p. 52.)
ACADEMY OF SCIENCE.]
XVIII.]
Appendix C.
CHEMICAL AND PHYSICAL CONDITION OF THE CUCARACHA, THE CHIEF SLIDING FORMATION.
By WARREN -J. MEAD and DONALD F.MACDONALD.
Page.
Analyses--------------------------------------------- ----------------------------I.............. 54
Analyses of ground water ------------*------------------------------- ---------------------------- 54
Discussion of the analyses ----------------------------------------------------------------------- 55
Chemical examination of rocks for colloids------------------------------------------------------ 55
Mineralogical composition of the Cucaracha formation -_ ---------------------------------------------- 58/
Physical properties of the Cucaracha rock-------------------------- *------------------------------- 60
Physical constitution of Cucaracha rocks ------------------------------------------------------- 60
Disintegration of dried rock by water --------------------------------------------------------- 61
Size of grain------------------------------------------------------------------------------- 61
Determination of the water contest of the Cucaracha formation----------------------------------------- 62
Introductory------------------------------------------------------------------------------- 62
Laboratory improvised ---------------------------------------------------------------------- 62
Results of experiments ---------------------------------------------------------------------- 62
Crushing strength of the Cucaracha rocks----------------------------------------- ------------------ 63
Table showing results of tests ----------------------------------------------------------------- 64
The nature of the slide movements, by Mr. MacDonald----------------------------------------------- 66
Table showing the slides and their yardages -------------------------------------------------------- -67
Figures 6-8. 5
80698-53-
[MEMOIRS NATIONAL
[VOL.
CHEMICAL AND PHYSICAL CONDITION OF THE CUCARACHA FORMATION.
By WARREN J. MEAD and DONALD F. MAcDONALD.
ANALYSES.
Most of the really big slides (see table, p. 67) and practically all of the sliding that occurred since 1914, developed in the Cucaracha formation. In order to answer the question as to why this formation was so prone to slide, investigations as to its physical and chemical character and its water content were made by us at the request of the committee of the Academy.
The formation, as already explained, consists of very fine-grained grayish-green, massive to locally bedded, slightly indurated, volcanic clay rocks of about andesitic composition, with two or more pebbly beds, some beds of coarse sandy friable tuffs and a few beds of lignite.
Analyses of the Cucaracha for mation.
ii II IV V Vi
Si02 --------------- 52.41 61.22 53. 13 48.36 50.44 48.83
A1202 -------------- 18. 18 13. 08 16. 65 18.98 17.82 13.63
Fe205-.------------- 5.20 4.58 5.19 10.53 9.84 5.49
FeO----------------- 3.67 .90 3.11 1.59 1.27 3.71
Mg0--------1.59 .85 1.80 .04 1.25 3.32
Ca~O----------1.62 1.99 2.34 1.84 1.41 436
Na21O---------------- .94 .50 .85 .61 .63 1.08
K0----------------- .81 .34 .59 .34 .43 .10
110--------------15.04 15.15 10.24 9.47 9.97 12.26
H20+..................................--- 4.74 6.71 5.90 5.41
Ti01----------------- .87 .85 1.04 .88 1.10 .95
C02------------------.11 .20 .43............ .10 .34
P205--------------- Trace ............. .06 .22 .04 .22
803 .......... ........... .......... .2..........12.... .....
Mn0----------------..20 Trace .05...... 02.....
100. 64 99. 66 100.34 99. 57 100.26 99. 70
The analysts of these samples are Dr. R. C. Wells for samples I, II, III, and V, and Dr. W. T. Schaller for IV and VI. The analyses were made in the laboratory of the United States Geologyical Survey
I. Collected from Gaillard Cut, near the southern edge of Gold Hill, at elevation 95 feet. Soft greenish massive volcanic clay rock, typical of the Cucaracha formation.
II. Same as I. Collected in Gaillard Cut 1,000 feet north of Gold Hill and about 6 feet below a lignitic shale bed in the Cucaracha formation.
III. Compound sample made up of about a score of samples of the typical greenish Cucaracha rock selected from various parts of Gaillard Cut.
IV. Soft red clay rock beds and lenses of which are found locally interbedded in the green clay rock of this formation.
V. Same type of rock as IV, but a compound sample made up of half a dozen samples selected from the red bed in -the Cucaracha formation as exposed in various parts of Gaillard Cut.
VI. Collected from tbe dark reddish, somewhat harder and better cemented zone that immediately overlies the relatively impervious lavabreccia flow in the Cucaracha formation. Not typical Cucaracha rock.
ANALYSES OF GROUND WATER.
Two samples of ground water, from the Cucaracha rocks, were collected, both from perforated iron pipes driven into the ground. Sample No. 1 was obtained from the undisturbed ground opposite station 1788, about 250 feet back from the east edge of Gaillard Cut, and about 14 feet below the surface. Sample No. 2 was taken from the slowly moving masses of the
54
ACADEMY OF SCIENCES.] CNIINO H UAAH 'RAIN xvill.]CODTO OFTECCRCAPR TIN
55
Ea 'st Culebra slide, 10 to 30 feet below the surface. These samples were analyzed by Mr. Jacobs, chemist, of the Ancon Health Laboratory, through the kindness of Maj. F. F. Russell, chief of -the laboratory, and Col. G. D. Doshon, superintendent of the hospital. The following results were obtained:
[Parts per million.]
No. 1. No. 2.
Total solids ........................... 210.0 331.0
Loss of ignition ........................ 52. 0 97. 0
Fixed residue------------------------....158.0 234. 0
Silicon dioxide------------------------... 13.2 15.4
Iron--------------------------------.... 1.1 10.3
Calcium-----------------------------.... 26.4 50.8
Magnesium -----------.-........-------.10. 4 16. 6
Potassium---------------------------.... 11.8 6.0
sodium-----------------------------.... 10. 8 14. 4
Sulphate radicle-----------------------... 9. 0 10. 2
Chlorine----------------------------....18.0 21.0
Bicarbonate radicle-------------------...126. 0 232. 4
Acidity to phenolphthalein:
No. 1, 100 c. c. required--------------------------------------------------- 1. 95 c. c. N1100 Alkali
No. 2, 100 c. c. required-------------------------------------------------- 11. 40 c. c. N1100 Alkali
DISCUSSION OF THE ANALYSES.
CHEMICAL EXAMINATION OF ROCKS FOR COLLOIDS.
The prevailing greenish color of the Cucaracha formation seems to be due to a fairly high content of ferrous iron (average, 3.11 per cent). Locally in the formation are many beds and. lenses at several different elevations, where the green coloration has been changed to red. These seem to represent old surfaces where deposition was held up for a while and atmospheric oxygen oxidized the ferrous iron to the ferric condition. This idea is supported by the field. relations as well as by the chemical analyses. The field relation show in many cases a gradual transition downward of the red beds into the greenish rock below with the red coloring following cracks and fractures into the green rock. Under the microscope iron oxide stain appears in some of the very minute cracks of the transitional material. No evidence of upward extension of this red coloration into the overlying green rocks was noted.
The chemical relations of the red beds and the normal green Cucaracha rocks is well shown by the "straight-line diagram."2 2 (Fig. 6.) In this diagram Sample 111, which is a composite average sample of the typical green Cucaracha rock, is compared with Sample V, which is a composite average sample of the red rock. The analyses of both these samples, calculated to a total of 100 per cent with the water given off below 1100 C. left out, are as follows:
Si 0 2.---- -------A1203 --- -------Fe20O--- ----- ----FeO..........--Fe............--Mgo...........--cao........---N a20 ----- ------K 20 ------------H 20 ----- ------H20O-.-.--------T i0 2 ----- ------P205.......---Co2.......----s...............-S0 2s-------------mno..........---
I
Recal. less
-H20-.
58. 97
18.48 5.76
3.45
6.72
2.00 2.60
.94 .65
5.26 1.15 .07
.48
v
Recal. less
-1120.
55. 80 19. 74 10.89 1.41
8.98
1.38 1.56 .70
. 48
6.54 1.22 05 .11
106 93.6 53
245
75
145 167
134 135
80.5
94. 4 140 433
.06 .02 300 .. . . . . .
100. 00 99. 90
2Straight-line diagram, Leith and Mead Metamorphic Geology, p. 288.
III
105 118.5 92 81.5
88
53 52 76 590
105 28 123
vi
Recal. less
-H20-.
55. 84 15.59 6.28
4.24
7.68
3.80
4.99 1.23 .11
6.19 1.09 25 39
100.006 ......*
-i-l-l-
56 PANAMA CANAL SLIDES-MEAD AND MACDONALD. [MEioIRs NATIONAL
The column headedII contains the quotients obtained by dividing the percentages in Column III by those in Column V. The figures in this column are then platted on the "straight line diagram" as follows:
5
$102 A/120T fe&20,2
Fe&C
Fe
mgJO
C6'O V 0 1<20
H20' 00,2
_______V________ Il _____ ___ __
00
$102
Fe R0 Fe0
GAI N LOSS
60____G 70 80 90 100 150 200 300 4-5 -/0-$o
C02
FIG. 6.-This diagram compares the normal greenish unoxidized Cucaracha rock, with the reddish oxidized phase of the Oncaracha.
The points platted in the above diagram represent for each constituent the number of grams of altered rock required to contain the amount of that constituent present in 100 grams of fresh rock. The green rock is considered as the fresh phase and the red rock as the altered phase. It seems to be fairly reasonable in cases of weathering to assume that alumina has remained practically constant. On the diagram all constituents whose points fall to the right of the alumina point (94), have been lost relative to alumina, and all with points f alling to the left of the alumina point have been relatively gained. If alumina has remained constant during the alteration, these losses and gains are, of course, absolute. The diagram shows that in changing from the green to the red phase SiO,, FeO, MgO, CaO, NaO, 1 12, and CO, were lost' and FeO,5, total Fe, and 1120 were gained with TiO. showing no appreciable change. In other words, TiO, has remained constant with alumina. These changes are in every respect typical of surface weathering. Oxidation of iron is shown by the increase of FeO3, and the loss of FeO. Absolute
Nahi20 1H20
_______II___ ii ii [ [ I I lH
Tio2
ACADEMY OF SCIENCES.]CODTO OFTE UAACA OR TIN XVIII.]CODTO OFTECCRCAFR TIN
57
increase of iron content is common in weathering because of the fact that FeO in the form of earbonate or other salts is readily carried in solution and later precipitated by oxidation. Organic material at the immediate surface may reduce iron to the ferrous form, after which it is carried downward and reprecipitated. Loss of MgO, CaO, NaO, and K,0 is, of course, in keeping with weathering, as these compounds form soluble salts. (See analyses of ground water, p. 54.) Increase in water (hydration) is typical, and loss of CO, is, of course, to be expected. TiO, like alumina is resistant to solution and it is interesting to note that it has remained constant with the alumina.
8102
Fe' m/450
/120
C02 TI70R
I I 11f)' .1 1 1-iI- i-
FIG. 7.-This diagram compares the slightly hardened and cemented phase with the normal Cucairacha rock.
0
8102
Fe2OJy Fe 0
0
CC7 0 Nc7 2 0' k(2 0
V20
C02
I UR
Sample VI represents a type of rock much better cemented. than the average Cucaracha material. The lava flow in the Cucaracha formation on which the bed represented by this sample rests, probably acted as -an almost complete barrier to downward moving ground water. The ground-water solutions mingling against this barrier deposited a part of their mineral content, thus cementing the contiguous rock. This hypothesis is supported by the chemical evidence. -Comparing Sample VI, calculated to a total of 100 per cent with the water given off below 1100 C. left out (see table, p. 58), with the normal Cucaracha rock of Sample III, by means of the "straight-line diagram" (fig. 7), we find that it has suffered a considerable
GAIN LOSS
0 6'0 70 80__30 /00 /50 200 300 +-5 -/0-50
______________ I _______ ......] ____ E~[ ________ [ ______ I _____ IIIIIiiIIL]'
t 1 r ~ t
___________ I _________ I _______ .1 ____ _______ .1. ______ 1 _____ ~
____I I__IJIII A__]~1IL[IIIII
t /.j r
- i
j
58
PANAMA CANAL SLIDES-MEAD AND MACDONALD. [MEORSNTOL
loss of K,0, but has gained greatly in all the other ingredients, except CO, and TiO, which have remained relatively constant. In other words, the bed represented by this sample seems to have been altered to its present phase chiefly by addition of minerals to the normal Cucaracha rock. It is an incipient type of anamorphism, through cementation.
The analyses of ground water from the Cucaracha rocks show that sample No. 1, which was taken from the undisturbed rock 250 feet from the edge of the Culebra Cut, contains much less material in solution than sample No. 2, which was taken from crushed material actually sliding. This is as would be expected, especially as water from the surface is kneaded into the moving material and, at least locally, gets under considerable pressure, thus enhancing its solving capacity. It also has a chance to be brought into contact with a larger surface area of solids than it would in the very slow circulation which it must have in the fine grained undisturbed rock, thus increasing its opportunity to take minerals into solution. By comparing the water analysis with the straight-line diagram it will be seen that No. 2 sample carries increased quantities of all the mineral substances which the altered (red) rock lost. It also carries more iron which the red rock gained. However, the iron gained by the red rock was probably to a considerable extent precipitated from solution.
MINERALOGICAL COMPOSITION OF THE CUCARACHA FORMATION.
As already explained the Cucaracha formation is made up chiefly of light-greenish, soft, very fine-grained rock, which has a somewhat soapy feel. At certain horizons this rock was oxidized to a reddish color. Thin sections of it were cut in the laboratory of the Geological Survey, after first boiling the rock chips in a mixture of Canada balsam and shellac, to overcome their friability.
Microscopic examination of the unoxidized material revealed a highly altered rock. In a cloudy groundmass some angular fragments of feldspar and a considerable number of quartz fragments 0.03 to 0.1 millimeter in size were noted. The feldspar seems to be largely plagioclase and shows more or less sericitization. Local aggregates of minute flakes appear to be muscovite or paragonite or both. A few small patches of green chiorite were noted, and other cloudy slightly greenish areas may be chlorite. The gray to slightly dark greenish, cloudy matrix appeared, in the main, to be kaolin, and this determination was later verified by the chemical data. Calcite was observed in some of the sections, as well as a few fragments which may be epidote. Minute grains having some resemblance to titanite were seen and a few black specks of magnetite. In general the microscopic determinations were difficult to make and are unsatisfactory owing to the cloudy and decomposed character of the minerals. In many of the sections certain of the larger grains (up to 0.5 millimeter) were observed to be slightly rounded and to consist of highly altered fragments of basic rock or basic mineral. Many of them are inclosed or partially inclosed in a thin film of translucent, slightly doubly refracting substance. In order to have this mineral determined, thin sections, and specimens of the rock were taken to Dr. H. E. Merwin, of the Geophysical Laboratory Carnegie Institution of Washington. Dr. Merwin's report is as follows:
In the thin sections of the green Cucaracha rock pale to colorless rims appear around certain grains of minerals or of altered rock fragments. The rims are either cracks or films of a scaly mineral, or both. The scales of the mineral are nearly parallel with the surfaces of the grains. During the drying of the rock fragments irregular shrinkage took place, and cracks opened along the surfaces of easy separation afforded by the films of scaly mineral. There are ramifications of films or irregular veins of this mineral, which is probably a light colored mica, in parts of the rock that are not distinctly granular. Also, there are cracks shown in the sections, which do not follow films of the platy mineral.
In the sections of the red Cucaracha rock granularity is only faintly preserved. These sections are much cracked, but there is no evidence that the cracks have been controlled by the distribution of any mineral.
It is interesting to note that according to Dr. Merwin's findings there were two periods of shrinkage in this rock. In the first, certain grains shrank more than the matrix rock which held them, thus leaving the shrunken grains partly or wholly surrounded by a minute opening. This opening subsequently became filled with the mineral substance described. Later, prob-
ACDBY F CINCS. CONDITION OF THE CUCARACHA FORMATION. 59
ably as the specimen had its content of ground water dried out, a second shrinkage occurred which resulted in fracturing along the easy parting of the old mineral filled cracks.
After microscopic examination the analyses of the rock were recalculated into the minerals observed, with due consideration given to the probable mineral relations and proportions which such an altered material might have. For this recalculation to a mineral basis the analyses were first recalculated to 100 per cent, with the water given off below 1100 C. left out. In the results which follow, zeolites are not taken into consideration, for their presence is not definitely known. In any case they can not be present in more than very subordinate amount, for the material necessary to their composition is in small quantity only, and much of it is necessary for the formation of other minerals known to be present. The table as shown is obviously not absolutely accurate, but it gives a fairly close approximation to the actual mineral composition of the rock.
Recalculation into minerals of Sample III, typical greeish Cucaracha rock.
Per
Mineral. cent- 5l02 A1203 Fe2O3 FeO MgO cao NaO K20- 1120- TiO Cos P20 MnO S3 age
58.97 18.48 5.76 3.45 2.00 2.60 0.94 0.65 5.26 1.15 0.48 0.07 0.06 0.13
Calcite........................ 1.00 ....... ....... ....... .......... 6....... ....... ....... ........4....... ....... .......
1.98 X
Albite ....................... 3.98 2.74 .77 ........... ............ .7....... ....... ....... ....... ....... ....... ........
56.23 17.71 .47
Anorthite ...................... 4.33 1.87 1.59 .................. .7....... ....... ....... ....... ....... ....... ....... .....
54.36 16.14 11
Muscovite..................... 5.52 2.50 2.13..... ....... ....... ....... ....... 65 .24........... ....... ....... .....
51.86 3.99 X 5.02
Chlorite....................... 5.52 1.80 .01 ...... ....... 2.00 ....... .......... 1....... ....... ....... ....... .......
50.06 12.98 X 4.31
Epidote....................... 1.28 .48 .30 .17 ....... ....... .30 ....... ....... 3....... ....... ....... ..........
49.58 12.68 5.59 .81 4.28
Paragonite ..................... 3.99 1.81 1.54 ........... ....... ........4........ 8....... ....... ....... ....... .......
47.77 11.14x 4.10
Titanite...................... 2.82 .86..... ....... ....... ....... .81 ...... ....... ....... 1.15 ........... ....... .....
46.91 x x
Kaolin ....................... 28.20 13. 13 11.14........... ....... ....... ....... ....... 3.94 ....... ....... ....... .........
33.78 x7 16
Magnetite ...................... 8.09 ....... ....... 5.59 2.51........... ....... ....... ....... ....... ....... ...... ..........
X .94
Melanterite .............................. ....... ....... .13 ............. ....... ....... ....... ....... ....... ....... ... .13
.81
Iron meta-silicate ............... 1.60 .73 ....... ....... .81 ............. ....... ....... ....... ....... ....... ... .06 ...
.33.05 x x Silica........................ 33.05 33.05 ............ ....... ....... ....... ....... ....... ....... ....... ....... ...............
X
Total .................... 99.64
The analysis given above has been recalculated to a total of 100 per cent after elimination of water given off below 10 C.
The percentage of the soapy minerals, kaolin, chlorite, muscovite and paragonite, shown in the above recalculation, is very large (43.24 per cent). Such a high percentage of greasy hydrous minerals undoubtedly added considerably to the mobility of the slide. As might be expected, this unoxidized sample contains no limonite.
The reddish Cucaracha material, under the microscope, shows, in general, a further degree of alteration than the unoxidized rock. The microscopic determination of its mineral content was not very satisfactory. However, in general, it seems to contain about the same minerals as the unoxidized rock with the addition of considerable limonite. The analysis of sample V recalculated into minerals, in the same manner as for the other analyses, shows the following percentages:
60 PANAMA CANAL SLIDES-MEAD AND MACDONALD. [MEORSNTONL
Recalculation into minerals of Sample V, typical red (iucaracha rock.
Per
Mineral. cent- SO2 A120 Fe2O FeO MgO CaO Na2O K20 120 TiO2 C2 P20 MnO age.
55.80 19.74 10.89 1.41 1.38 1.56 0. 70 0.48 6.54 1.22 0.11t 05 0.02
Calcite .............................. 0.25........... ....... ....... ........1....... ....... ....... ........11....... .......
1.42 X
Albite .............................. 2.96 2.04 .57..... ....... ....... ....... .35 ............. ....... ....... .........
53.76 19.17 .35
Anorthite ........................... 2.10 .91 .77 ............. .......4....... ....... ....... ........... ....... .......
52.85 1.40 1.00
Muscovite............................ 4.08 1.85 1.57..... ....... ....... ...... ....... .48 .18 ........... ....... .......
51.00 16.83 X 6.36
Chlorite.............................. 3.80 1.24 .70....... ....... 1.38 ........... ........8....... ....... ....... .......
49.76 16.13 X 5.88
Epidote............................. .68 .25 .17 .08 ....... ....... .16 ....... ....... .02 ........... ....... .......
49.51 15.96 10.81 .84 5.6
Paragonite........................... 2.97 1.35 1.14 ...... ....... ....... ....... .35........ .13 ........... ....... .......
48. 16 14.82 X 5.73
Titanite ............................ 2.98 .92..... ....... ....... ........ .84 ...... ....... ....... 1.22 ............ .......
47.24 x x
Kaolin............................. 37.40 17.50 14.82........... ....... ....... ....... ....... 5.22 ........... ....... .......
29.74 X .51
Iron meta-silicate ...................... .91 .41 .............. ........ ....... ....... ....... ....... ....... ....... .........
29.33 .91
Magnetite........................... 2.93 ....... ....... 2.02 .91 ............ ....... ....... ....... ....... ....... ........
X
Limonite............................ 9.30 ....... ....... 8.79 ....... ....... ....... ......... ........ ....... ....... .......
Silica........................... 2933293...........3 9.3 ....... ....... ....... ....... ....... ....... ....... ......... -------- --Total .......................... 99.69 x
The analysis given above has been recalculated to a total of 100 per cent after elimination of water given off below 110 C.
The above recalculation, like the other, is, of course, approximate only but cannot be very wide of the mark. It shows that limonite is present in considerable quantity, that magnetite is relatively scarce, that melanterite is absent and that kaolin is much more plentiful than in the unoxidized rock. The percentage of slippery hydrous minerals, kaolin, chlorite, muscovite and paragonite, in this oxidized sample is 48.25 to which might be added 9.30 per cent of limonite, making a total of 57.55 per cent. Such a very high content of slippery, hydrous minerals must have added very greatly to the mobility of this material, especially after it began to slide.
PHYSICAL PROPERTIES OF THE CUCARACHA ROCK.
PHYSICAL CONSTITUTION OF CUCARACHA ROCKS.
The typical Cucaracha rock in its original saturated condition is readily mashed in water into a more or less plastic, claylike material, without subjecting it to anything in the nature of fine grinding. If this claylike mass is further washed with wAater, it yields a residue of sand consisting of more or less angular quartz grains and grains of magnetite, many of the latter showing crystal faces. The claylike material is derived from the matrix, and from thoroughly altered mineral fragments. If a dried specimen of the rock is examined under a binocular microscope, small grains of quartz and magnetite and rounded grains of light and dark greenish minerals are observed in a soft matrix which has a horny or waxy appearance and texture, and in which no individual grains can be seen even under considerable magnification. The whole rock shows much evidence of shrinkage due to drying, as there is a myriad of minute cracks extending in every direction. The majority of grains on being tested with a needle point are as soft as the matrix material and very closely resemble it in texture and appearance. The only hard grains observed are magnetite and quartz, although other unaltered minerals are undoubtedly present in minute amounts.
ACAieM O CINBI. CONDITION OF THE CUCARACHA FORMATION. 61
DISINTEGRATION OF DRIED ROCK BY WATER.
When a sample of Cucuracha rock, which has been dried at room temperature in an atmosphere of comparatively low humidity or at 1000 C. in a drying oven, is placed in water it disintegrates very rapidly, yielding a fine-grained mud. This disintegration is so intimate as to practically separate the individual grains, and the sand grains are then easily separated by washing without further crushing. This phenomenon accounts for the ease with which mud flows are developed on the dried devegetated surfaces of the slides by the action of heavy rains at the beginning of the rainy season.
SIZE OF GRAIN.
If the undried rock is mashed to a paste and shaken up in water, a considerable proportion remains in suspension for a long time. Only a part of this suspension is held on the finest filter papers. After filtering through double thickness of Schlucher and Schull No. 597 filter paper and double thickness of hardened filter paper No. 575 of the same make, the resulting suspension was of about the translucency of milk.
On settling in a 2-inch tube, a clear slightly opalescent zone appears at the top bounded below by a definite upper surface of the slowly settling material. This surface moves downward at the rate of about 0.05 inch per day. The slightly opalescent zone at the top does not clear by settling, and examination with the ultra microscope shows that it is very rich in colloidal material.
In order to compare the composition of the very fine material in suspension with the entire rock, a sample of typical fine-grained rock was divided into three parts, one part was reserved for analysis an d the other two parts were reduced to pulp and shaken up in water. Of the resulting suspensions one was dried at the end of 24 hours and the other at the end of 38 days. Both stood in 100 cubic centimeter burette tubes about 30 inches long. After standing for 38 days 100 cubic centimeter of fluid contained 0.26 gram of suspended matter dried at 1000 C. Following are the analyses of the three samples, dried at 1000 C.
A is a sample of entire rock, B is material in suspension at end of 24 hours, C is material in suspension at end of 38 days. A', B', C' are the same analyses on a water-free basis. A", B"/, C" are the same analyses recalculated to the same percentage of A1, as the entire rock in order to afford a basis for comparison.
Analyses of Cucaracha rock and water suspensions of the same.
A B C A' B'W IC A"l C"
Entire Sus- us Recalculated with water Recalculated to same per rok pended pended omtecn 10asA
rok24 hours. 38 days. oitd cnAlOas.
Si02 ---------------------------- .. 59.67 54. 11 51. 1 63.65 57.20 57.30 63.65 45.40 45.45
A1203 ----------------------------------------- 17. 98 22.90 21.6 19. 20 24. 20 24.20 19. 20 19.20 19.20
Fe2,Os ----------------------------------------- 5.62 7.08 1111 6.00 7.48 112.45 6.00 5.93 1988
FeO------------------- ----- ---- --- ------ 3.11 3.24 (1) 3.3 3.43 (1) {10.5 (9)
MgO ----------------------------------------- 2.78 3.25 2.5 2.96 3.44 2.50 2.96 2.73 2.78
CaO----------------------------------------- 2.35 2.23 2.0 2.51 2.36 2.24 2.51 1.87 1.78
Na20O---------- -------------------------------- .83 .35 .4 .88 .37 .45 .88 .29 .36
K20. O --_----------------- ---- ------- .61 .62 .5 .65 .65 .56 .65 .52 .44
R120 ----------------------- ---------- ------------11.1 ................. .......... .......... .......... ....
H 20 --- --- ---- --- --- --- -- --- -6--36--5.19---.....6.3.......L...9.... ... ....... ...... .... ...... ... .... ... ...... .... .. .....
V0O2---------------------------------------..... .80 .80 (1) .85 .85 (1) .85 .67 (1)
100. 11 99.97 100.3 100.02 99. 98 100.00............ ..........
1lTi02 and FcO included with Fe203These analyses show that the material in suspension is essentially the same in composition as the entire rock except in SiO, which represents the quartz grains which settle out rapidly.
Drying the rock causes it to lose in some measure the property of remaining long in suspension in water. In order to make a quantitative comparison of the undried and dried rocks two samples of equal weight were taken from a single specimen of typical Cucaracha rock, in its natural undried condition. Those samples each weighed 40 grams. One was crushed to
632 PANAMA CANAL SLJDES-MIEAD ANID MACDONALD. [MEORS VTOAL
a paste in a mortar and shaken up with enough water to make 100 cubic centimeters in a 2-inch tube and allowed to stand 24 hours. The fluid was then decanted and dried at 1000 C. The other sample was dried at 1000 C. for 24 hours, then crushed and suspended in exactly the same manner as the first sample. The dried sample weighed 34.45 grams. The suspension from the undried sample after 24 hours amounted to 2.564 grams, and from the dried sample 0.937 gram, both dried at 100' C. These weights are equivalent to 7.5 and 2.7 per cent, respectively, of the weight of the original sample. This seems to indicate that the nature of the material is changed by drying, possibly by the dehydration of colloidal material.
DETERMINATION OF THE WATER CONTENT OF THE CUCARACHA ROCKS.
INTRODUCTORY.
Engineers and geologists who have given serious study to land-slide problems in most cases have found ground water to be a large factor in promoting the movement, and thorough underground drainage to be the most effective remedy. This is probably the reason why several eminent men advocated that the canal authorities try extensive underground drainage to stop the slides.
With many eminent men insisting that drainage be at least tried, it became necessary either to spend a great deal of money in putting in tunnels and wells to test the drainage idea, or to make some inexpensive tests that would give positive information on the drainage question. In conference with President Van use, chairman of the committee from the National Academy of Sciences, it was agreed that experiments of the latter type could and should be carried out as soon as possible.
Efforts were first directed toward measuring the free moisture content of the typical Cucaracha rocks from samples collected from below the water level and in well-drained portions of the Cucaracha above the water level, by determining drying losses and apparent porosity.
LABORATORY IMPROVISED.
For drying the specimens a dry room about 7 by 10 feet, where a temperature of 1000 C. could be maintained, was provided. Air-tight cans for taking 8 to 15 pound samples of the rock and conserving its moisture content until the samples could be brought to the laboratory and weighed were made. Laboratory space for determining the specific gravity of the samples, both by weighing in air and in water and by the pycnometer method, was arranged for.
RESULTS OF THE EXPERIMENTS.
When all was ready 21 representative samples of the Cucaracha or sliding formation were taken from below the water level of the canal. The average of these saturated samples contained 12.20 per cent of water by weight, equivalent to 27.9 per cent by volume. The average of 19 samples taken from well above the level of ground water, where the rocks were much jointed and fractured, and therefore perfectly drained, contained 11 per cent of water by weight. As shown above, 12.2 per cent of water by weight fills all of the pore space of the rock; therefore 11 per cent by weight fills only 90.6 per cent, leaving 9.4 per cent of the total pore space as having been emptied by drainage and by drying. Now, 9.4 per cent of 27.9 per cent is 2.6 per cent of the total volume of the rock. This shows that natural drainage of the most perfect. kind would not remove more than 9.4 per cent of the water by weight, equivalent to 2.6 per cent of the volume of the rock. However, most of the samples from the drained rock were taken very close to the suirf ace, so that very likely they lost some of their water through drying out by the heat of the sun, for the dry season was more than a month old at the time they were collected.
These f acts show that while the sliding rocks have a high percentage of pore space the pores are mostly of subcapillary size and are filled with water which adheres to the walls by molecular attraction and can not therefore be drained off. These experiments the writers believe to have established that all cures by drainage which had been offered to and urged on the canal authorities were practically futile.
ACDBY FSENE.] CONDITION OF THE CUCARACHA FORMATION. 63
CRUSHING STRENGTH OF THE CUCARACHA ROCKS.3
The average specific gravity of 43 samples of the Cucaracha rocks (see p. 64), without taking account of the pore space, is 2.757. But each cubic foot of the rock is made up of 72.1 per cent solid mineral and 27.9 per cent of pore space filled with water (except in the narrow zone above ground water, which may be neglected for the purposes of this calculation).- Therefore, each cubic foot of rock weighs 141.6 pounds. The following table shows the gravity pressure on the base of any perpendicular column of the rock a foot square and 100 to 500 feet high:
Pressure Pressure
per square per square
foot. inch.
Pounds. Pounds.
100 feet high........................ 14,160 98. 3
200 f eet high----------------------...28,320 196.6
250 f eet high----------------------...35, 400 246.0
300 feet high ----------------------- 42,480 294.9
350 f eet high----------------------...49,)560 344.0
400 feet high----------------------...56,640 393. 2
450 feet high----------------------...63,720 442. 5
500 feet high----------------------...70,800 491. 0
The weakest sample tested of the typical Cucaracha rocks crushed at 41,184 pounds per square foot, which pressure is attained at a depth of 290.8 feet.
The strongest sample tested of the typical Cucaracha rocks crushed at 70,272 pounds per square foot, which pressure is attained at a depth of 496.2 feet.
It is noteworthy that with relatively light pressure most of the blocks under test began to exude water. In some cases air bubbles were also pressed out of minute fissures.
Two very important weakening factors active in large masses of this rock could not be shown or measured by the testing machine; they are, time and the larger fissures, faults, and cracks which intersect the formation. It is impossible to determine how much should be deducted from the crushing-strength tests of these rocks in order to show the actual sustaining power of very large masses of the rock, because of such weakening factors. If 50 per cent be deducted then the pressure attained at a depth of 146 feet below the surface would be greater than the sustaining power of the rock. However, it is certain that the weakest sample tested did not represent the weakest bed or zone in the Cucaracha rocks exposed in the slopes of Culebra Cut. The strength of any steep bank under strain is, in general, no greater than the strength of the weakest bed or zone in it. From the foregoing it will be obvious that, in the main, banks or slopes in the Cucaracha formation which approach the perpendicular would be in some danger of failure if more than 100 feet high. The break or deformation type of slide had very little development until the cut was more than 125 feet deep.
3 The Bureau of Standards, Washington, D. C., made crushing strength tests on seven samples of the Cucaracha formation, and the results of those tests are used in the above calculations.
64F 'PANAMA CANAL SLIDES-MEAD AND MACDONALD. [MVNTOL Table showing results of moisture tests made on the Cucaracha formation.
Weight Loss on drying.
Weight of _______No. of of spei- Seii spedi- specs- men Per graymen. men after cent ity.1 as col- drying Weight. of unlected. at dried 1000 C. specimen.
Ounces. Ounces. Ounces.
1 140 129 11 7.8 2.76 2.- 126 .112 14 11.1 2.82 3a..- 100 88 12 12.0 2.87 3b.-- 28.5 24.5 4 14.0 2.79 4..... 144 129 15 10.5 2.83 5-. 108.5 95 13.5 12 5 2.76 6.- 152 134 18 11.9 2.73 7..- 124 107 17 13.6 82.77 8..- 172 151 21 12.2 2.77 9 -.... 149 129 20 13.4 2.77 10-- 126 108 18 14.25 2.74 11..-.. 96 82 14 14.60 2.74 12.. 203.5 184 19.5 9.6 2.77 13-. 117 104 13 11.0 2.77 14. 96 84.5 11.5 12.0 2.77 15-. 86 76 10 11.6 2.77 16.. 102 88 14 13.6 2.78
17.... 116 101.5 14.5 12.5 2.77 18-. 128 110 18 14.0 2.73 19-- 120 104 16 13.3 2.76
20.... 96 81 15 15.6 2.77 21.... 110 97 13 11.8 2.78
22.. 120 104 16 13.25 2.80 23.... 136 120 16 11.6 2.76
24.... 129.5 116 13.5 10.5 2.75 2. -.. 171 148 23 13:5 2:76 26.. 134 116 18 13.5 2.73 27-. 165 141 24 14.5 2.78 28-. 192 174 18 9.3 2.69 29-. 119 104 15 12.5 2.71 30a .- 103.5 99 4.5 4.4 2.68 30b... 128.5 117.5 11 8.5 2.69 30c- 131.5 117.5 14 10.6 2.69 31.... 162 146 16 9.8 2.69 32- 94 80 14 14.9 2.70 33-... 147.5 136 11.5 7.8 2.69 34.... 156.5 137.5 19 12.1 2.77 35-. 186.5 167 19.5 10.5 2.75 36.-. 157 140.5 16.5 10.5 2.74 37.. 125 109.5 15.5 12.4 2.75 38-. 154.5 140 14.5 9.4 2.73 39-. 112 101 11 9.8 2.75 40.. 125 114 11 -8.8 2.82
41-.. 133 120 13 9.7 2.77 44-. 191 17.5 16 8.4 2.78 45..- 155 141.5 14.5 9.3 2.76 46.- 174 159 15 8.6 2.83 47.- 140 125 15 10.6 2.825 48.... 138 120 12 8.6 2.71 49.... 121 104 17 14.0 2.61 50- 112 98 14 7.5 2.78 42-- 142.5 132 10.5 7.4 2.80 43- 110.5 99 11.5 10.5 2.81
Average4 ..................................
Loss on drying in terms of per cent
of volume of specimen
Abnormal speciSatu- Dand mens. rated sDraie speci- mes mnmens. Satu- Drained.
rated.
28.0 31. 2
24.7 28.2 26.8 30.6 27.8
26. 7 30. 3
28.3 30. 8 29. 8
33.7 26.7
30.0
27.8
22.6
20.2
20. 8
30.0
227 25.7
26.4 24.3
30. 1
2.5
24. 3 21.4
219 22. 8 21.3
2.9
20.4
18.2
23.9 32. 2
18. 1
18. 7
31. 4 31.9
21. 4 19. 8 22.7 18.7
2.7 18.2
27.91 25.3j1 28.0 1 27. 9
Description.
Fine-grained basic argillite, 6 inches from surface. ine-grained basic argillite, 12 inches from surface. Fine-gralned basic argillite.
Do.
Do.
Do.
Do.
Do.
Fine-grained basic argillite, 6 inches from surface. Fine-grained soft dark shale.
Do.
Fine-grained basic argillilte, 6 inches from surface. Fine-grained basic argillite, 1 inch from surface. Fine-grained basic argillite.
Fine-grained basic argillite, slightly reddish,
sheared.
Fine-grained basic argillite, reddish. Fine-grained basic argillite, slightly sandy. Fine-grained basic argillite, between shale beds 6
inches apart.
yine-grained basic argillite, slightly lignitic. Fine-grained basic argillite, considerable red
coloring.
Fine-grained basic argillite, fairly firm asnd hard. Fine-grained basic argillite, above andesite flow,
slightly hardened.
Fine-gralned basic argillilte.
Do.
Fine-grained basic argilite, sandy. Fine-grained basic argillite. Sand and fine gravel bed. Fine-grained basic argillite. Sand and fine gravel bed, 6 inches from surface. Sand and fine gravel bed, 12 inches from surface. Sand and fine gravel bed, saturated. Sand and fine gravel bed.
Do.
Do.
Fine-grained basic argillite. Fine-grained basic argillite, red bed. Fine-grained basic argillite.
Do.
Do.
Do.
Fine-grained basic argillite, red bed (not completely drained). Fine-grained basic argillite.
Do.
Do.
Fine-grained basic argillite, red bed. Fine-grained basic argillite, red beds. Fine-grained basic argillilte. Fine-grained lignitic shale. Fine-grained basic argillite. Gold Hill meta-tuff .
Do.
1Determinations by pycnometer methods.
wx 100
2Formula used for tbis calculation is P= R
where P is percentage of pore apace; W is weight (in ounces) of absorbed water; 1B is weight of dry dock (in ounces); and G is tbe mineral specific gravity of the rock. Leith and Mead, Metamorpbic Geology, p. 285.
sSpecific gravity by pycnometer of the specimens 7 to 15 inclusive and 17 was not obtained, so the average specific gravity for tbe same type of rock is substituted.
4 The above substitution of values bas slightly changed the general results published in the Annual Report for 1916 of the Governor of the Panama Canal., p. 601. The present results show that only 2.6 per cent by volume of the water content of the rocks might be drained oil as compared with
3.6 per cent by volume mentioned in the above report, taken from data which is believed to be less accurate than that here presented.
ACADEMY OF SCIENCES.]CODTO OFTE UAR HA OMTIN XVIII.] CNIINO H UAAH OIUIN
Watern
27.9 percel7t or ~th& vo/4.lr,7&of- /7e roc,4-
A ott f tt .141A /7?/1/71 pO55'/jh }he cdrcz.'1c/ o7F 2.6 ,e'e~OP zt/he
Ro k
72. / pcr c e r
FIG. 8
65
436 PANAMA CANAL SLIDES-MEAD AND MACDONALD. [MEMOIRS NATIONALT
The accompanying diagram (fig. 8) is drawn to scale. If the volume of a block of the average Cucaracha rock be taken as 100 per cent, then 27.9 per cent of that volume is pore space, which is filled with water when the rock is saturated, and 2.6 per cent of it, or 9.3 per cent of the total water content, might possibly be removed by very perfect drainage, by putting down a drainage well, say, every square foot.
The average moisture content of- Per cent.
19 drained specimens was by weight........................................................... 11. 2
21 saturated specimens was, by weight----------------------------------------------------........12.0
19 drained specimens was by volume----------------------------------------------------........25.3
21 saturated specimens was by volume---------------------------------------------------.........27.9
THE NATURE OF THE SLIDE MOVEMENTS.
By DONALD F. MACDONALD.
The nature of the movements which gave rise to slides, of the deformation type is, the writer has always maintained, rooted in two mechanical principles--deformation by rupture and deformation by plastic flow. Rupture takes place where the rock is d eformed without being under heavy pressure from all sides. The tests made in the laboratory of the Bureau of Standards, because of their nature, necessarily showed deformation by rupture with only a suggestion of plastic flow. The material moving in the upper zone of the slides is deformed by rupture. However, the writer believes that the deformation in the lower zones of the slides takes place by plastic flow. He thinks that in the beginning stages of deformation, the zone of plastic flow is relatively deep. As time and movement go on, however, ground water is kneaded into the mass. This water enters not only from cracks leading from the surface, but also from the contained ground water in the rock being pressed into all the cracks and open-' ings that form through deformation (such pressing out of water happened when blocks of the rock were put under pressure in the testing machine). With extra water thus kneaded into the mass the zone of plastic flow must rise very considerably and the mobility of the sliding mass must be greatly increased. Such considerations as these, together with other lesser reasons, caused the writer earnestly to advocate in all his reports (see Ann. Rept. Isth. Can. Coin. for 1912, pp. 208-209, and 1913, p. 580). lessening of the slopes where high and steep, by terracing with steam shovels so as to prevent initial deformative movement.
Various data concerning slides of Gaillard Cut.
Formation and location.
Las Cascadas agglomerate:
East Bas Obispo................
East Haut Obispo.............--West Buenavista..............--East Buenavista..............---East Las Cascadas.............--East Whitehouse..............--West Whitehouse.............--West Whitehouse yard.........--East Powderhouse.............--East North La Pita............--East Lower La Pita............--East Upper La Pita.............
West Cunette...................
East Empire ...................
West division office, Empire... Culebra:
West Lirio .....................
West New Culebra Viliage......
East Hagan slide................
Cucaracha:
West Hodges Hill...............
West Culebra, Zion Hill........--East Culebra Gold Hill..........
East Cucaracha.................
West Contractors Hill, north.
West Contractors Hill, south.
East Cucaracha Viliage ..........
East Paraiso....................
Culebra:
East Pedro Miguel.............----
Date when slide first developed.
September, 1910 ... September, 1908...November, 1908...May, 1912.... February, 1908 .... October, 1908.. May, 1914 .... June 1912. October, 1909 .. September, 1912...May, 1910 .... December, 1909... September, 1910...May, 1912 .... May, 1910 ....
Aprli, 1912.--September, 1909... February, -1913 ....
August, 1912.. October, 1907-.. January, 1907. July, 1905.---January, 1907.-July, 1908.---September, 1911...March, 1907 ...
January, 1913 --.
Total.....................................--------
Date when slide became quiescent.
Cubic yards excavated to date.
1909 1to July 1, 1911.
I* I'
Ar 1112... Jun,1913 ... ActiveJuy., 1916.. AuguLst., 1913.. March, 1912 ... October, 1911- May, 1914 ....
June1913 ...
August,'1913.-December,. 1909...March, 1912.--February, 1911....M ay, 1913.---April, 1914.--1914...........--1914...........--1915...........--1916...........--1916...........--June, 1915.---Quiescent .... June, 1911.---Dccember, 1912... May, 1912 ....
May, 1913.----
111,000
18,064 43, 301 286, 000
20, 000 63,)000
2,..329,.784 2,--722,-164
211,9 036
July 1, July 1, July 1, July 1, July 1,.
1912. 1913. 1914.2 1915.3 1916.
117,000 117,000........... ............ 117,000
18,064 18, 064 ............ ............ 18,000
162, 000 162,1000........... ............ 262, 238
....503,.000 48,000 ------------ ------------ 48,000
50;00 503,)000.......................-- 503, 000
286,2000 509, 000.......................-- 509, 000
.......... 1000 ------------ ------------ 45,000
~413, 000 543,000.......................-- 614,260
.... 181, 100.......................-- 189,600
30,1000 30,1000 ............ ............ 54, 733
20, 000 20, 000 ............ ............ 20, 000
67,000 67, 000 ............ ............ 67, 000
i ... 933,700.--..........--.I.--------.1,071, 272
210,000 258,000-----------------------.. 260,415
--------- 221,200-- ----------- ------------33,6
33,020,987
61765)000O 8)872600........ .........
4, 290, 000 5,966,200............
2, 890, 000 3, 859, 500 ............ 4,351,526 9,901,602
216, 000... 221,...000 .... ...- .. .... ... .... ... ...
57,1000 231,000........... ............ 231,000
385,000 385,000............385,000
......... 3,300---------....-3,300
. ------123, 009, 6641 29, 524, 217 1 36, 077, 585
4 47, 8R80, 475
Cubic yards Cubic yards Area in remainig remaining ares. Jan. 1, 1916.1 July 1, 1916."
'455,0001
'355,000
114, 195, 000 1 7, 801, 269
500,000 ............
............
-----------------------
500, 000
2.8 .6
5. 1
1.2 11. 5 6. 5
1.0 5.8 1. 7 3. 1 1.7 .9
20. 0 2.6
3.2 1 700 19.64 2,250 11.16....
'11.68
60. 8 70. 5
60. 4
5.7
.2
5Thirteen slides were in motion during fiscal year 1908-9, and from them 884,530 cubic yards of materials were removed and 993 cubic yards were estimated to be remaining in motion.
2 From July 1, 1913, to July 1, 1914, 6,514,553 cubic yards were removed. Cbcyrs
8'By dredges..................................................................... 6,361,450
Total from slides July 1, 1914, to July 1, 1915 ................................-----6, 553, 368
4 In addition to this quantity, dredges have removed from Cut 1,270,459 cubic yards, mostly due to small slides.
6 Total approximaate.
Frontage.jI
Feet.
500 1,00
Length 0 east.
Feet.
600 2
300
1,400 1,9800
8y 00
525 300 (
Y25
12 00
1,100
600
23,555
1,7100 2, 900
600
-1 1 1-1- 1-
............ 1- 1. 112600 1
ACADEMY OF ScIunNcISsj XVIII.]
Appendix D.
MECHANICS OF THE PANAMA CANAL SLIDES.
By GEORGE F. BECKER.
CONTENTS.
Page.
O observations on ethe islides ...... ........... ............ ........... ........... ........... ......... 69
Limiting depth of disturbance------------------------------------------------------.............. '70
Conditions in a wide cut ------------------------------------------------------------------------ 70
L im iting values ofof .k .......... ........... ........... ........... ........... ........... .......... 72
Examples of slide curves ----------------------------------------------------------------------- 73
H ydrostatic anaanalogy ......... ......... .......... .......... ......... .......... .......... ......... 74
Form ation f of pruptures ....... .......... ......... .......... .......... .......... .......... ......... 74
Bulging of canal bottom ------------------------------------------ft------------------------------ 74
Effect of the form of the banks--------------------------------- -------------------------------- 75
N ote oonffinitesstrains................ .................... ................... ................. 76
Figures 9-11
80902--
[MIMOIRts NATIONAL
[VoL.
MECHANICS OF THE PANAMA CANAL SLIDES.1
By GEORGE F BECKER.
OBSERVATIONS ON THE SLIDES.
Early in 1913, before water was admitted, I spent some weeks in examining the geology of the Culebra Cut, now officially known as the Gaillard Cut, with special reference to the origin of the landslides.2 These appear to me to be of two kinds-mere superficial slips on joint planes or other slippery surfaces and deeper-seated "breaks" as they are known by the engineers. It is only with the latter that this paper is concerned.
The breaks in their inception are marked on comparatively level banks by groups of cracks or narrow fissures nearly parallel to the cut, and these almost immediately develop into series of step faults with small throws, many of them only a fraction of an inch in height, the hade where not vertical being invariably toward the canal so far as I could observe.1l Many of the steps of these faults are only a yard or two in width. There seems little order in the time of formation of the cracks; in some breaks groups of small faults first appear rather close to the cut, those at a greater distance from it developing later. In others the earliest cracks are hundreds of feet from the canal and the intermediate ground splits up afterward. In all the breaks which I could examine the first small movements involved no perceptible gaping, or none of the same order of magnitude as the throws of the faults. At or about the same time as the cracks on the bank were formed nearly horizontal cracks also appeared in the cut near the bottom of the bank, but which of these were the earlier it seemed impossible to decide.
After a break has made a fair start the cracks more remote from the cut gape and show underlying curved surfaces which reach the general level of the top of the bank nearly at right angles or crop out almost vertically, and at the outcrop the vertical cross section of such a surface shows a very moderate radius of curvature. The surfaces of rupture are fairly smooth, many of them slickensided a little below the outcrop, but not smooth enough to make accurate measurements of their radii of curvature practicable. As nearly as I could discover these radii measured between 100 and 200 feet. Where these underlying surfaces are exposed to a considerable extent it is apparent that the radii of curvature increase rapidly with increasing depth, and some exposures from which disintegrated material had been removed appeared to prove that as the cut is approached the radius of curvature becomes very large indeed.
Movement of the slides perhaps never entirely ceases, but it varies greatly in velocity, from a fraction of an inch a day to many yards. After considerable motion has taken place the sheets of'rock are broken up and the external surface of the slide becomes as rough as a choppy sea.
A certain amount of consolidation and of what might be called secondary cohesion sometimes occurs in a slowly moving slide of large dimensions after the material has been reduced
I Published as Professional Paper No. 98-N of the U. S. Geological Survey.
2 1 had the great advantage of Mr. Donald MacDonald's companionship throughout these field studies.
8 Mr. MacDonald records that "some of the blocks sank a little in front and tilted up in the rear, so that they were a yard above the front part of the block behind." This behavior was unusual, and I saw no instances of it. Local inhomogeneities in the bank might perhaps bring about irregularities in surfaces of rupture which would account for exceptional throws of 2 or 3 feet. No other suggestion on this subject occurs to me.
69
70 PANAMA CANAL SLIDES-BECKER. [MEORSNTOL
to a chaotic condition. In such cases well-developed curved surfaces of rupture and step faults form, indistinguishable in general character from the initial disturbances in the solid bank. This surprising fact indicates that definite mechanical laws of wide applicability underlie the formation of slides. I was witness to these phenomena in the Cucaracha slideY and they have made their appearance in other and more recent breaks.
During the progress of a large slide upheaval of the bottom of the canal may take place from time to time, showing, that deformation of the rocks extends to a certain depth below the deepest excavation; but this upheaval does not attend every spasm of activity in the slide, nor does the amount of material thrust up indicate that deformation extends more than a few yards beneath the bottom of the canal. A layer of rock say a hundred feet in width, buckled by nearly horizontal pressure, would show, even if it were only a couple of yards in thickness, mounds of rubble as much as 20 or 30 feet in height, or of the order of magnitude of the observed upthrusts.
LIMITING DEPTH OF DISTURBANCE.
To simplify the mechanical problem as much as possible, suppose the case of a level plain underlain to a great depth by an ideally homogeneous rock. At any depth in this rock the pressure will be hydrostatic and equal to the depth multiplied by the density. Suppose a narrow trench to be sunk vertically in this rock, the width being so small that caving of the sides can be prevented by mine timbering. Then, because of the one-sided relief of pressure there will be at the bottom of the cut a horizontal stress, directed from the wall into the cut, which is equal to the product of the depth and the density. This stress will tend to produce a horizontal shear and to drive the bottom of the wall into the cut. If the cut is sunk deep enough, so deep that the stress is equal to the resistance of the rock to simple shearing stress at the elastic limit, this deformation will occur and the wall will bulge.
This seems a rather hasty statement, but in the last section of this paper the strains are considered in detail; it is there shown that the elastic limit for simple shear would be reached long before the limit for mere linear compression, and that of all elementary resistances that resistance which opposes stress such as is exerted by a pair of scissors is the weakest.
Let the limiting depth at which this one species of flow makes its appearance be denoted by y, so that if p is the density the hydrostatic pressure is py,, which is also the value of the shearing stress.
CONDITIONS IN A WIDE CUT.
The hypothesis of a narrow timbered cut was employed in finding the limiting depth, y, in order to avoid the complication of a caving bank. Let a wide cut be substituted, one a mile wide if the reader chooses, but let the bank be vertical. Then even before the depth y, is attained, any real rock wall would break down or cave. But imagine for a moment the rock replaced by a substance so tough that, though it would undergo permanent deformation at the same limit as the rock, it would hang on long enough to be studied. A ductile substance, such as wrought iron, would act in this way.
Consider a surface of uniform deformation nearly as deep as y, and extending into the wall. This surface will surely not be horizontal, for such a strain would imply the expenditure of an infinite amount of energy.
Before caving can take place in a homogeneous bank the material of the bank must be strained to its elastic limit. The vertical cross section of the bank must therefore include a line along which the strain is -uniform. This line must reach the top of the bank somewhere, and it may be assumed that the line is curved, because that is a far more general hypothesis than that it is straight, besides being in harmony with observation.
In figure 9 OBC represents the bank and ABOD a part of the cut. The x axis, or OX, is taken at a depth y, from the original surface, and EC is a curved line along which the shearing stress is uniform. The problem is to find its equation.
ACDEY F CINCS. MECHANICS OF THE PANAMA CANALj SLIDES. 71
At any point the original hydrostatic pressure was (y1 y) p, but excavation of the cut having disturbed the original equilibrium and brought about strain reaching the elastic limit, ha's developed a shearing stress which is equal to (y, y) p per unit length and which is of itself in-75 adequate to cause flow. But there is another manifestation of stress to be considered. The shear- A ing stress is equivalent to a tension in the direction ----of the curve, say T per unit length. Let V/' be the y angle which the tangent to the curve makes at xy; let 64, be an elementary angle and bs a corresponding arc. Then elementary mechanics shows that the tension, T, acting along the arc &s is equivalent to a normal pressure T6V.4
It has already been explained that py, per T6o
unit length is the shearing stress needful to strain FIG. 9.-Curve Of uniform tangential strain. the mass to its elastic limit for simple shear. Hence if stress of this intensity is to be set up along the curve EC the following equation must hold good:
T8W&+ (y y) p~s =y,pebs
or, more briefly,T
TP~
Here &164.I, is the radius of curvature, say R, while T/p is a constant characteristic of the material and essentially positive. It may therefore be replaced by b2 and then Ry = b2, which is the most general equation of the elastic curve.
Replacing R by its value in terms of dy/dx and d2y/d X2 and integrating once gives
Y 2=-2b2 COSiV----------------------(1
where C is a constant of integration. The form of the curve depends on the value of C. For the present problem it is evident that the curve can not cross the. x axis and that y can not become negative, so that C must equal or exceed 2b It is easily proved that if C= 2b2 the equation represents a curve coinciding with -the x axis for an infinite distance. This is not a case to be considered, and therefore C> 2b 2. The equation then represents the elastic curve of Euler's eighth class, a diagram of which is given in Thomson and Tait's Natural philosophy,' S611, figure 7.
For some purposes equation (1) is convenient enough. Thus if the ordinate of the point at which the tangent of the curve is vertical is called y, then C= yv2; while if the ordinate at the point where the tangent is horizontal is y., then y.2Iy,2-, 2b2. But values of the abscisswe are not so simple.
It is needless to say that the geometry of the elastic curve has been thoroughly known for a century and that this is no place to expound the subject. A few results, however, must be set down. By substituting & 2p r 0 =b 2
where p is a variable angle and kc is the sine of a constant angle, it will be found that
2b1~k2 sin 2
Then also dx= cot 4dy= cot 2sdy .. .. .. .. .... .... ..(2a)
4 See Tait, P. G., Properties of matter, p. 253, 1894; or Lamb ,H., Statics, p. 276, 1912.
72 PANAMA CANAL SLIDES-BECKER. [MVNTOL
and x takes the form of an elliptic integral.
For purely practical reasons (the scope of tables of elliptic integrals) it is convenient to reckon x negatively, or to the left of the origin in figure 9. Then, in the conventional terminology of these integrals,'
if, for example, the horizontal distance from the origin to the vertical tangent of the curve is required, 4' = 7r/2 and p = r/4, and the valueof x can be computed from tables of elliptic integrals. Because of symmetry, positive values of x have the same absolute value as negative values.6
The element of area of the curve, ydx, is independent of k:
ydx = 452 (i dp =52 cos ~
so that
f ydx=b2 sin 4 .. .. .. .. .... .. ... .....(4)
a result which can be found directly from (1) and (2a).
The length of the curve counted from the horizontal point is given by
8
k.............k.................................. (5)
and s/S is thus simply proportional to the first part of the value for -xIS.
It should be remarked that 5 2 is an absolute constant dependent only on the density and tenacity of the rock, so that geometrically b is the unit in which lengths are computed and 2 the unit area. On the other hand, kc varies from curve to curve of a family of curves, all of which share a common value of b, but as ic is the sine of an angle it can not exceed unity.
LIMITING VALUES OF k.
It has already been pointed out that if 0=25. the elastic curve is a horizontal straight line coinciding with the x axis. The same equality implies that kc is unity, and therefore, for the problem ufider discussion,'ik must always be the sine of an angle lest than sij2. It is equally evident that ic can not vanish, for were it to do so the curve would intercept the vertical axis at an infinite distance. There are other reasons for supposing that ic can not be very small, and these can be very briefly stated. In this discussion it has not been needful to consider any strains except those at the elastic limit, but the general theory of elastic strains shows that at the edge of a vertical cliff or bank there will be no strain at all, and for some distance from such an edge the strains will be exceedingly small. Hence strains reaching the elastic limit are not to be considered near this edge. It might be possible, but it would not be worth while, to determine how near to this edge the elastic limit could be reached.
On the other hand, it is very important to consider how far back a curve of critical shear can reach, and this I believe to be a simple problem. From the manner in which the equation of the elastic curve was derived it is apparent that the pressure due to tension is a secondary phenomenon due to elastic strain. It is unthinkable that this part of the pressure should exceed the whole pressure requisite to produce flow. But when the curve crops out on the bank at 90' to the horizontal, the pressure due to tension at the outcrop exactly equals the critical tension, y~p. Hence for a given value of yi the lowest possible curve is that which intersects the level bank at right angles. From this condition the maximum value of Ic can be determined.
5For the meaning of the symbols in equation (3), see for sample Peirce's "Short table of integrals."1
6 Equation (3) is substantially identical with that given by Lamb (Statics, p. 279), who, however, takes the origin at a different point, making x and ~ p disappear together, so that the y axis includes the maximum value of y. In (3) the origin is so transposed that x and b disappear together, so that, as required for the problem in baud, the y axis passes through the minimum value of y, or the point for which v =?r2
ACXV YOFSCFNCI.] MECHANICS OF THE PANAMA CANAL SLIDES. 73
EXAMPLES OF SLIDE CURVES.
In order to illustrate conditions resembling, to a first approximation, those met with in the Gaillard Cut, I have computed a few values of the more important elements of the curves, and these are tabulated below. It is easy to see that only relatively large values of k- = sin a are of interest, and I have begun with a = 75'. Taking x, as the abscissa of the vertical tangent, it is found from equation (3), while if y, is the value of the ordinate for x = 0, y0,/b = 2 cot a. Then y12/b2 = Y0,2/b2 + 2. The fundamental relation y/b = bIR makes it easy to find the radii of curvature answering to x~y, and x,,y0. For the purpose of the diagram it is not requisite to compute other points; after describing an arc at the axis of symmetry with R0/Ib and a second arc at x~y, with R1/b, the two can be connected without serious error by the help of a curved ruler.
Points on the elastic curve.
To estimate an appropriate value for b, it is requisite to adopt some value for the resistance of the rock either to shearing stress or to crushing. The ultimate resistance to crushing of such materials as soft-burned brick, inferior 1 concrete, and the poorest sandstones is somewhat less than 3,000 pounds per square inch. 3 The Cucaracha formation is probably of similar strength, and I will assume its resistance to be 2,760 pounds. In the concluding section of this paper reasons are given for supposing that toe62 times the resistance to shear is about equal to the resistance to crushing, and this implies-0 ----- ---- I.
that for the Cucaracha the resistance to shear is 325 pounds per square inch. This is the FIG. 10.-Elastic curve foray 750, 800, 850, 900. weight of a column of rock of a density 2.5 times that of water and 300 feet high.
If the curve for which a = 89' is selected and y, is taken as 300 feet,
2
b Yi Cot2=44,972; b=212
2 (1-2 cta)
By multiplying all the lengths given in the table by 212 a consistent set of values is obtained.
In the diagram (fig. 10) the height of the bank above the x axis is taken as 300 feet and the curve for a 890 cuts it perpendicularly at a distance of 841 f eet from the y axis. The curves for smaller values of a give larger values for y, and therefore cut the 300-foot level at acute, angles.
According to the theory here set forth, a limit is set to the vertical height of a cliff of any rock. From results obtained by the United States Geological Survey it appears that granites show resistances up to 34,000 pounds per square inch, which would correspond to a cliff 3,700 feet high. The brow of El Capitan, in the Yosemite Valley, stands 3,100 feet above the valley, but the top of the dome, some 2,000 feet back from the brow, is about 500 feet higher.
k xi/b yo/b yr/b Ro/b RLJb
sin 75 ............. 1.3411 0.5358 1. 512 1.866 0.661
sin 8----- -1.7094 .3526 1.458 2.836 .686
sin 850- --......2.3728 .1750 1.425 5.714 .702
sin 890 .....3.9690 .0350 1. 415 28. 570 707 sin90............. 00 0 1.414 00 .707
74 PANAMA CANAL SLIDES-BECKER. [MEORSNTONL
HYDROSTATIC ANALOGY.
If two rectangular blocks of very clean glass are placed in a dish, parallel to one another, and if water is added until the faces of the blocks nearest together are wet to the top in consequence of capillarity, then the vertical cross section of the water surface between the blocks is the elastic curve represented by equations (2) and (3); the height of the blocks above the general water level is given by y~b, and the amount of water raised above this level by capillarity or surface tension is P2 per unit length for each wall of the channel between the blocks. If the surface tension is T and the density is p then T/p = 2.
This system is in stable equilibrium, the surface of the water is minimal for the boundary conditions, and, as the equilibrium is stable, the gravitational potential is a minimum. The whole system may be supposed solidified without disturbance of equilibrium. One-half of this model, to the right or the left of the point at which the capillary curve is lowest, represents the mass beneath a slide on the Gaillard Cut. The whole model represents the slide surfaces as they would be were the cut extremely narrow, provided that the material sliding in were removed as fast as it came until the slides "died."
This very perfect analogy and the theory of this paper seem to me to show that the profile of the bed or bottom of a straight watercourse or river, flowing through a homogeneous stretch of country, must tend to approach the elastic curve, and that this profile is also most suitable for a canal.
FORMATION OF RUPTURES.
Thus far the discussion has been limited to conditions appropriate to incipient flow; the rock has been supposed strained to its elastic limit, but short of the point of rupture. In such materials as rocks, which are to be classified as brittle substances, the difference of stress between the so-called limit of solidity and the breaking point is extremely small. Moreover, as a matter of course, real rocks are not homogeneous.
Suppose that the limit of solidity has been exceeded by a minute stress increment, but that along some small arc of the elastic curve the rock were more brittle than elsewhere; then evidently a local crack would develop; the resistance along the entire curve would be diminished pro tanto; the stress on the remaining larger portion of the curve would be correspondingly increased; further rupture would follow; and, as it appears to me, the crack would extend from one end of the curve to the other in much less time than is required to state this conclusion. So, on a frozen lake, when a sudden fall of temperature occurs, a crack starts with a report at some point along the shore and tears, booming, across the ice sheet at a velocity approaching that of sound.
If before rupture there is plastic flow along a given curve, then after rupture the overlying mass can move by gravity; for till rupture occurred motion was opposed by cohesion, and when this is overcome resistance is diminished. Thus there is a surplus of energyy available to accomplish work.
BULGING OF CANAL BOTTOM.
The necessary and sufficient condition f or flow is that Ry = 6, and the smallest value which Rt can reach is R, = b2/y. The stresses which bring about this condition are due to the tendency of the cliff to settle down into the cut, and this tendency will persist until flow takes place along the basal curve, for which 4,i = 7r/2 at the outcrop.
Strain can not be confined to levels above the bottom of the cut, for the moment the bank begins to sag, even within the elastic limit, adjoining masses are stressed to some extent, and these stresses must extend, with diminished intensity, to great distances. Thus even while the cut is shallow there must be elastic strains along the basal curve. As the depth of the cut increases the strain along, this curve must increase until it approaches the elastic limit, both in the wall and below the cut in the plane of the wall.
Now, suppose that the cut is nearly but not quite down to the basal curve and that, by some local inequality in the resistance of the material on some part of the basal curve, or in consequence of some jar, due perhaps to movements in the bank, a short local crack forms on
ACADEMY OF SCIENCES.] MECHANICS OF THE PANAMA CANAL SLIDES. XVIlII.
75
some part of the basal curve; then the question arises whether or not this crack will spread. Movement of the mass overlying the curve will be opposed by the horizontal resistance to crushing or buckling of the mass underlying the floor of the cut and extending down to the curve, but when this stratum has been reduced to a very small thickness the crack may extend and cross the vertical, thus splitting off a layer of rock immediately beneath the cut. As has been pointed out above, the formation of a crack along the curve suddenly releases an amount of the energy of position of the bank corresponding to the cohesion which existed until the crack formed and spread. At the expense of this energy buckling or bulging of a thin bottom layer may take place.
This seems to me an adequate qualitative explanation, of the upheavals of the floor of the cut observed during the later part of the excavation. That shock had something to do with these upheavals is suggested by the fact that continuous slow upheavals corresponding to the slower movements of the slides were not observed. Upheavals accompanied only the spasmodic accelerations of slide movement. This phenomenon is a harbinger of what would occur if the cut were extended down to the full depth y,for then the bottom and sides of the cut would ooze in continuously by plastic flow.
EFFECT OF THE FORM OF THE BANKS.
To simplify discussion it has been assumed that the canal was a vertical cut through a flat country underlain by homogeneous rock, and of course these assumptions are not in accord with the facts. But the country is rather flat; and as the underlying rock is a solid mass, though not a strong one, the variability of load near'the surface must be fairly well distributed at depths of more than 100 feet.
Until slides began to give trouble the banks of the cut were very steep-quite too steep, in fact, as everyone would now concede. It is well to consider what would have been the effect of giving the excavation lower slopes.
The ordinary theory of earth pressures on retaining walls is based on the existence of an angle of rest in a pile of discrete particles. It appears to me to be totally inapplicable to conditions in the Cucaracha formation, for the mere existence of breaks demonstrates that the mass possesses continuity. The rocks of the Gaillard Cut behave very much as a mass of agaragar jelly might do if a rectangular mold of this substance, a foot or so in depth, were turned out on a horizontal table. If the jelly were of the right degree of stiffness, the edges of the mass would first sag, and then breaks would make their appearance; but nothing resembling a constant angle of rest would be developed. To work out a complete theory of the relief of pressure in such a jelly, or in the Cucaracha formation, due to an inclination of the walls, would probably be very difficult. Nevertheless, very simple considerations show that sloping the walls is an effectual method of reducing the pressure.
If the Gaillard Cut were replaced by an exceedingly strong wall, the pressure against the wall would be hydrostatic. For a small change of depth the increment of pressure would be p (y, y y y), and the whole horizontal pressure from the surface to depth y, 6y" would be e-(y yo)'.
Now, imagine a plane inclined to the horizon at 450 and passing through the point X = 0) Y = Yo- This plane would cut off a triangular slab, say of unit thickness and of mass
The amount of frictional resistance depends primarily upon normal pressure, so that if F is the frictional resistance and N the normal pressure
F
-X=Atanv
where it is the coefficient of sliding friction and v the angle of friction. Now, F can not exceed Sthe normal pressure N, which excites it, so that 1, can not exceed unity and v can not exceed 45". Hence friction can not prevent movement on a slope of 45'. Thus if the triangular mass
76 PANAMA CANAL SLIDES-BECKER. [MEORSNTONAL
of rock (or of jelly) were actually separated from the remainder of the mass, friction would not prevent it slipping down the steep slope. The tangential pressure which the mass w would exert on the 450 plane would be W!-V/2, and this would be resolved into a vertical pressure and a horizontal pressure each equal to w/2.
Thus of the whole hydrostatic horizontal thrust exerted against the vertical wall, just one-half is exerted by the triangular slab. Hence also sloping the bank of a cut at 450 would diminish the horizontal thrust to one-half of its maximum value.
It is not difficult to perceive by the further application of elementary statics that the thrust would be still more diminished by making the slope smaller than 45'.
The precaution of giving the banks a low slope might have prevented the occurrence of slides, but as a remedial measure, after break s have developed to a considerable extent, it seems to me of little avail. After the basal curve has developed into a crack, the material overlying it is either in motion or in unstable equilibrium; and sooner or later all, or nearly all of it, will reach the bottom. Slides of origin similar to those of the Gaillard Cut are by no means confined to the Canal Zone. In my opinion banks of cuts should be watched with extreme care, and the moment any cracks make their appearance all other work should be suspended until a safe slope has been established. Breaks should be prevented, because they can not be cured.
NOTE ON FINITE STRAINS.
Plastic flow is continued deformation without change of density. It takes place at the so-called limit of solidity. During flow, therefore, a solid must, be treated as compressed to a constant extent, and as the elasticity of volume is perfect, when stress is relieved the original volume is restored. In nearly all cases a solid mass undergoing flow is to be treated as incompressible.
This limit of solidity depends on the type of strain to which the mass is subjected and to some extent on viscosity. It would also depend on heterotropy, but this paper deals only with isotropic matter.
In any strain ellipsoid there are two symmetrically oriented sets of planes of maximum tangential strain or maximum slide. If the strain is a rotational one (so that the groups of material particles through which the ellipsoidal axes pass vary with the progress of the strain) then there is a difference in behavior of the mass on these two sets of planes. Along that set of geometrical planes which rotates more rapidly through the mass, or on which the material particles change more quickly, greater resistance is offered to flow or rupture than on the other set, because the resistance to be overcome is rigidity plus viscosity and because viscosity offers great resistance to a sudden stress but very small resistance to a stress slowly applied.
In one strain, called simple shear, shearing motion, slide, or scission by various writers, there is one set of these planes which is fixed relatively to the material, while the other set of planes of maximum tangential strain changes its position relatively to the material particles more rapidly than in any other strain.
In scission, therefore, flow will be more easily produced on the fixed set of planes than in a strain of any other type; but on the other set of planes flow will be less easily produced in scission than in a strain of any other type. Scission is due to a couple acting against a resistance. It is the only strain produced in a rod of circular cross section when the rod is twisted about its axis.
Irrotational or pure shear, usually denoted simply as shear, is the simplest conceivable deformation. If a sphere is so distorted that one diameter retains its length unaltered while two other orthogonal diameters, in a plane perpendicular to the first, are so changed that their product is constant, and if all three diameters pass through the same material particles at all stages of the strain, then this strain is a shear, though not a simple shear but yet far simpler than a simple shear. Both strains are illustrated in figure 11.
A pure shear may be conceived as the resultant of two scissions whose rotations are equal and opposite, a fact of which use may be made in the present discussion.