Engineering Geology - Elsevier Publishing Company, Amsterdam - Printed in The Netherlands

RECENT QUICK-CLAYSTUDIES, 6

Q U I C K CLAYS OF E A S T E R N C A N A D A

CARL B. CRAWFORD

National Research Council of Canada, Division of Building Research, Ottawa, Ont. (Canada)

(Received December 11, 1967)

SUMMARY

This paper describes the results of research carried out during the past
fifteen years on the quick clays of eastern Canada. The mineral composition
and structure of these clays is similar to that of other glacial marine clays but they
often exist with an unusually low salt content in the pore water. There is evidence
of strong rigid bonds between particles of those deposits that are overconsolidated.
The bonds are broken when the soil is loaded and this may lead to large volume
changes or to exceptional loss of strength if drainage is not permitted. Some studies
have been made on the physico-chemical characteristics of the soil and on its
fabric but most of the research to date has involved laboratory and field obser-
vations of the response of the soil structure to stress changes. The important
influence of rate of loading has revealed the need of careful interpretation of
laboratory tests for field application.

INTRODUCTION

"Quick clay" is a class of natural soils with unusual characteristics. The

most striking feature of a quick clay is its tendency to transform from a relatively
brittle material to a liquid mass when it is disturbed. The engineering impli-
cations of dealing with such a material are readily apparent. In Canada, quick
clays occur commonly in the valleys of the St. Lawrence and Ottawa Rivers,
and in recent years a considerable amount of geological and engineering research
has been devoted to them.

GEOLOGY

Toward the close of the last glacial period, the northern part of North
America was depressed several hundred feet. When the ice finally retreated, the
present St. Lawrence River valley was invaded by marine waters. The extent of
this arm of the ocean called the Champlain Sea is outlined on Fig. 1.

Eng. Geol., 2(4) (1968) 239-265

240 c.B. CRAWFORD

80 ° 7~ ° 77 ° 68 ° 64.°°__.-..

5~

'

,/ I

x---.~r,a SEA G ~ - ~ " " q ~--/6~

oNrAP/O ~J) k~ % v,. ?/

Fig.1. The Champlain Sea.

The Champlain Sea existed for about 4,000 years, from 12,000 years B.P.
to 8,000 years B.P. (KARROW, 1961). During that period the land rose by more
than 700 ft. (230 m) by isostatic adjustment and the eustatic rise in sea level was
more than 100 ft. (33 m) (KENNEY, 1964). The highest recorded marine level is
690 ft. (226 m) above the present mean sea level. This is near Ottawa, where the
maximum sea depth has been estimated at several hundred feet (JOHNSTON, 1917).
In a geological review of the Champlain Sea, KARROW (1961) described
the origin of the sediments that accumulated in it. In the early stages these were
derived primarily from stream erosion of land underlying or adjacent to the re-
treating ice. Deltas were built where the streams entered the sea leaving typical
deposits of variable material, the finer sediments being carried farthest from shore.
Underwater slumping created further variations in the final sedimentary deposits.
During later stages, as the land level rose isostatically, some of the original marine
deposits were eroded from the shorelines and redeposited in the sea.
Recent geological work in the Ottawa area suggests possible major re-
deposition of marine clays into fresh water (GADD, 1962). It is reasoned that the
salinity of the water changed rapidly due to a sudden influx of fresh water (prob-
ably from the Great Lakes region) accompanied by major erosion. This opinion
is based on the occurrence of large volumes of oxidized non-calcareous and non-
fossiliferous clays lying unconformably on unoxidized, calcareous, and fossiliferous
clays. This would explain the low salt content in the Ottawa clays (CRAWFORD
and EDEN, 1965); however, if these upper clays were actually deposited in fresh
water, some explanation must be found for their open structure and sensitivity.

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 241

The structure may be due to flocculation caused by residual cations remaining

even after washing by the fresh water.
The sensitive fine-grained sediments of the Champlain Sea are called
"Leda" clays in the engineering literature, a name said to have been first used by
the early Canadian geologist Dr. J. W. Dawson.

COMPOSITION OF LEDA CLAY

Few samples of Leda clay have been subjected to mineral analysis. X-ray
diffraction studies on a dozen samples from the St. Lawrence valley revealed
chlorite, illite, amphibole, quartz, and feldspar in all samples (KARROW, 1961).
BRYDON and PATRY (1961) carried out extensive tests, including mechanical,
chemical, and mineralogical analyses, on 18 samples from four sites in the Ottawa
area, and found no marked differences in mineralogy. Quartz, feldspars, amphi-
boles, micas, and chlorites were found in fractions of all sizes. Most of the samples
were slightly calcareous, the coarser material containing the higher amount of
carbonate. The origin of the clays was attributed to the igneous and metamorphic
rocks of the Canadian Shield. In a study of three samples of marine clay, ALLVN
and JOHNS (1960) reported hydrous micas, chlorite or vermiculite as the abundant
clay minerals.

STRUCTURE

The "structure" of a clay soil is a complex space frame of solid particles

held together by plastic and rigid bonds. The "fabric" refers to the size, shape and
geometric arrangement of particles, ranging from a parallel to a random or card-
house orientation. Quick clays have a structure that is "sensitive" to distortion
and their structural properties change dramatically when they are strained.
The Leda clays, like other sensitive clays, are thought to have a card-house
fabric because of their environment during deposition. Particle dimensions range
from about 50m bt to about 50 p, or by a factor of about 1000. The coarser fractions
are probably inert materials of more or less spherical shape and the finer fractions
are the elongated or plate-shaped clay minerals. The clays are generally "inactive"
(EDEN and CRAWFORD, 1957).
Several undisturbed specimens of Leda clay have been viewed with a scanning
electron microscope (GILLOTT, 1968). Water was removed from the specimens by
freeze drying and the viewed surface was cleaved on a chilled metal block by a
sharp blow from a chilled chisel while the specimen was frozen. Typical micro-
graphs, magnified approximately 4500 times, are shown in Fig.2. Fig.2A shows
the components of a clay soil (62 ~ clay size) which has a natural water content
of 58 ~ and a liquid limit of 53 ~. It has a measured preconsolidation pressure

Eng, Geol., 2(4) (1968) 239-265

242 c.B. CRAWFORD

of 2 tons/sq, ft. (kg/cmZ). Fig.2B shows a clay from the same general area but
with a preconsolidation pressure of 5 tons/sq.ft. (kg/cm~), a natural water content
of 55 %, and a liquid limit of 31%. Both micrographs are of horizontal planes and
similar views on vertical planes suggest a random orientation of particles in
Fig.2A and a more parallel arrangement in Fig.2B. This difference is anticipated
from the differences in overconsolidation.
Efforts have been made to measure the change in fabric of Leda clay during
anisotropic consolidation using an X-ray diffraction technique (QuIGLEY and
THOMPSON, 1966). The technique is based on the principle that if the clay platelets
are oriented along a particular plane, the intensity of basal X-ray reflections
obtained by X-raying this plane will be greater than that on other planes. Undis-
turbed specimens were subjected to various vertical pressures up to 64 tons/sq, ft.
(kg/cm2), removed from the oedometer and impregnated with a polyethylene
glycol. A smooth horizontal surface was ground and each specimen was subjected
to X-ray analysis. The degree of parallelism was inferred from measurements
of the amplitude of the illite peaks obtained on the smooth surfaces.
The results of these tests showed little or no particle reorientation at pres-
sures up to the preconsolidation pressure. Beyond the preconsolidation pressure
the structure begins to break down and there is an abrupt increase in particle
parallelism. Similar tests on remoulded soil showed greater parallelism and lower
void ratios at all pressures. QUIGLEYand THOMPSON (1966) suggest that with
most of the bonds broken the particles are free to rotate and therefore assume
lower void ratios and greater parallel orientation under a given pressure. The
relationship between fabric and void ratio was the same for undisturbed and
remoulded soil. This led to the conclusion that the final void ratio under a given
load is a function of the degree of induced parallelism.
Fabric can also be determined by measuring the thermal conduction of a
clay specimen on various axes because the conductivity of the clay minerals is
highly anisotropic. If the mineral particles are parallel the conductivity in the
direction of parallelism will be greater than it is in a perpendicular direction.
Tests on natural Leda clay gave values of conductivity in the horizontal direction
only about 10 % higher than that in the vertical direction, compared with values
up to 70 % higher for a lacustrine clay, which is known to have a parallel fabric
(PENNER, 1963b). Consolidated remoulded Leda clay showed greater parallelism
than the natural soil. Penner also showed the differences in fabric of Leda clay
and the lacustrine clay by optical viewing of thin sections between crossed po-
laroids.

Fig.2. Electron micrographs of Leda clay (x4,500).

A. Clay soil with a natural water content of 58%, a liquid limit of 53% and a precon-
solidation pressure of 2 tons/sq.ft.
B. Clay soil with a natural water content of 55yo, a liquid limit of 31~ and a precon-
solidation pressure of 5 tons/sq.ft.

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 243

lOp

I(

lO~J
244 c.B. CRAWFORD

Shrinkage characteristics of Leda clay and the same lacustrine clay de-
scribed above have been studied by WARKENTIN and BOZOZUK (1961). The un-
disturbed Leda clay had about equal horizontal and vertical shrinkage, but for
the lacustrine clay the vertical shrinkage was three times the horizontal shrinkage.
This is attributed to the random orientation of Leda clay particles and the tendency
to parallelism in the lacustrine clay.
It is obvious from the open structure of Leda clay and from its stress-
deformation characteristics that there are often substantial intrinsic stresses or
bonds holding the particles together. The exact nature of these bonds is not
known, but they are considered to be of two types--rigid and plastic (CRAWFORD,
1963a). The rigid bonds appear to be especially important in the overconsolidated,

Fig.3. Demonstration of undisturbed and remoulded strength of Leda clay. (CRAWFORD,

1963a.)

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 245

high strength deposits. These may be due. to recrystallization between particles

or to cementation by such agents as iron oxide or carbonates. In the few chemical
analyses available, up to 7 ~ iron by weight (BURN, 1966) and more than 1 2 ~
carbonate (BRYDON and PATRY, 1961) have been found.

SENSITIVITY

The "sensitivity" of a soil is the ratio of its undisturbed strength to its

remoulded strength at the natural water content. Clays with a sensitivity greater
than 32 were termed "very quick clays" by BJERRUM(1954). The Leda clays usually
have a sensitivity greater than 20. The undisturbed strength is often several hundred
times as great as the remoulded strength. Very high values of sensitivity are not
reliable because of the difficulty of measuring the remoulded strength of very
soft (semi-liquid) clay (EDEN and KUBOTA, 1961). The relative strengths of a
typical sample of Leda clay are shown in Fig.3.
Early work in Norway showed a correlation between sensitivity and pore
water salt concentration but this relationship is not found in the Leda clay. In
the Ottawa area, for example, most of the clays have less than 2 g/l salt con-
centration, but the sensitivity varies widely (CRAWEORD and EDEN, 1965). In
order to investigate this apparent anomaly, PENNER (1963a) studied the influence
on shear strength of changing the electrolyte content of the pore water and found
that a slight increase could greatly increase the sensitivity. In one test the addition
of l g of sodium metaphosphate per 100 g of soil increased the sensitivity from
30 to about 2000. The effect was attributed to an increased repulsion between
adjacent particles caused by an increase in the electrokinetic potential of the system,
and this led to an experimental study of the electrokinetic behaviour of the soil
(PENNER, 1965).
Specifically, the electrokinetic potential (i.e., the potential difference across
the diffuse layer in a clay-water system) was computed from measured values of
osmotic flow when an external electrical potential was applied across a plug of
undisturbed soil. The variation of sensitivity with electrokinetic potential, shown
in Fig.4, is a clear indication of their interrelationship. The two inconsistent
specimens were of low plasticity and had low surface areas. Measurements of the
specific conductivities of the pore fluid (i.e., salt concentration) showed that soils
with high salt contents cannot achieve high sensitivities because of the flocculating
effect, but that at low salt contents the sensitivity may be high or low.
Penner's work demonstrated that sensitivity is not necessarily a function
of the pore water salt content but is related to the electrokinetic potential and
hence is consistent with the theory of interparticle repulsion predicted by ROSEN-
QVlST(1955). There is therefore no physico-chemical contradiction to the geological
view that some of the Leda clays were deposited in fresh or brackish water.

Eng. Geol., 2(4) (1968) 239-265

246 c.B. CRAWFORD

1000
p I I r [ r I I [ I I/'7--q
900

800

700

t~~ 600
>"

.5O0

3O0

k~

-.+--t--r~'l-@~"r i I r I r r , i i I r r i I
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 N 30 31 32 33 34 35
ELECTROKINETIC POIENIIAL M.V.

Fig.4. Electrokinetic potentials of undisturbed Leda clay as a function of sensitivity

(PENNER, 1965).

COMPRESSIBILITY

Because of its rigid structure, Leda clay has a low index of recompression,
but when the load-carrying capacity of the structure is exceeded the compression
index is very high. This results in a sharply defined "preconsolidation pressure"
in laboratory tests.
In some areas Leda clay is almost normally consolidated, but there are also
widespread deposits of substantially overconsolidated clays that can support large
loads without detrimental settlements. Reliable determination of the precon-
solidation pressure from laboratory specimens is therefore of great importance
and has prompted much research on the compressibility characteristics of Leda
clay.
For ordinary consolidation testing, the total load applied to the specimen
is doubled every day in order to get a reasonable spread of test points on a semi-
logarithmic plot. In a highly sensitive compressible clay, the change in void-
ratio between the last point on the recompression section and the first point
on the virgin compression section of the curve may be quite large. This presents

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 247

a special problem in interpretation of the preconsolidation pressure due to the

steepness of the virgin curve.
In a study of this problem a series of tests was made in which the load
increment ratio was varied between 1 and 1/10. It was found that with small
load increments the pressure/void ratio curve could be defined more accurately
than in normal testing (HAMILTON and CRAWFORD, 1959). It was noted that at
pressures near the preconsolidation pressure, the rate of compression was not
simply a function of pore pressure dissipation but appeared to be related to the
breakdown of the soil structure. In these tests, therefore, the load increments
were added when compression under the previous load had decreased to a rate
equal to that at the end of primary compression in normal tests. As a result there
was not much variation in the over-all rate of testing.
In a later series of tests, the rate of loading was found to have a substantial
effect on the pressure/void ratio curve (CRAWFORD, ! 964). Results of tests on three
specimens are shown in Fig.5. The lower curve is for a test in which load increments
were applied once a week. The middle curve shows the results for daily loading.
The upper curve was obtained by adding loads at the end of primary consolidation
under the preceding load. The first test lasted for 5 weeks, the second for 5 days,
and the third for about 3 hours. All tests were repeated at least once; the three
curves are typical of other similar tests. Three specimens were loaded under
a controlled rate of strain at about the same average rate as that in the third test

I Ill n=la I I I I I I II
i vC I - . ~ p c = l . 8 . ,
| ! pc-3.o
,~.~ ~
|~.~. ~-Load increments
% ~ at end of primary

~~ (3 hour test)

).2
tJ

o-10
z
Daily load
increments \ ~ \
(5 day test, ~ ~
uJ
ct.
:E
o20 _ \ \\
Week, y,0a0 A, N \ \N N
increments J
(5 weektest) "N'~"

30 L illl = = I I I Ilil~l~'._
1 lo
EFFECTIVESTRESS(KO /CM 2)
Fig.5. Compression-log pressure curves for various rates of incremental loading.

Eng. Geol., 2(4) (1968) 239-265

248 C. B. CRAWFORD

8
I 1 I I I I I
7
e-a

t~
36
c~

~5
~o
o~'~"-e.~Coniinuous loadincj

7-4
~,,.o__f"-"l ncremental loading

_a

z3
O

I I I I I I I
10 20 30 40 50 60 70 80
DURATIONOFTEST(DAYS)
Fig.6. Influence of test duration on estimated preconsolidation pressure.

in Fig.5 and they had equivalent effective stress/void ratio curves. The results
show that the preconsolidation pressures vary from 1.3 to 3.0 tons/sq.ft. (kg/cm 2)
depending on the method of test.
A further series of tests was made on more overconsolidated specimens
of Leda clay and preconsolidation pressures ranged from 4.9 to 7.4 tons/sq.ft.
(kg/cm2) (CRAWFORO,1965).
Research was then extended to much slower rates of loading with tests
lasting from 2 days to 3 months, and the preconsolidation pressures varied from
4.3 in the slow test to 5.3 tons/sq.ft. (kg/cm 2) in the fast test (JARRETT,1967).
These tests covered a wide enough range of rates to isolate a pressure zone in
which the structural effects of the soil were controlling the rate of compression.
In this zone, from approximately4 to 7 tons/sq.ft. (kg/cm2), about 4 times as much
compression occurred under the slow loading as under rapid loading.
From all of these tests it may be concluded that the observed variations in
the pressure/void ratio curves are functions of the test procedure. Specimens have
to be loaded in large increments in order to assess the hydrodynamic effects needed
to compute a coefficient of consolidation, but such loading of a thin specimen may
cause it to compress a million times faster than it would under field loading. The
tests have shown that even where hydrodynamic effects are negligible, the amount
of compression is highly dependent on the rate of loading, and the classical
concepts of consolidation as applied to highly sensitive clays require review.

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 249

In the Seventh Rankine Lecture BJERRUM(1967) assessed both the laboratory

and the field behaviour of a very uniform normally-consolidated clay and concluded
that after consolidation on a geological time scale " . . . a cohesive clay will develop
increased strength and a reserve resistance against further compression". This
"reserve resistance" allows the geological preconsolidation pressure to be exceeded,
under relatively short term loads, without significant volume change. This can
therefore account for the high values measured in the laboratory.
It is now possible to compare the influence of test duration (or rate of
loading) on the measured preconsolidation pressure. This is done in Fig.6 based
on tests reported by CRAWFORD (1965) and JARRETT (1967). The solid curve is for
tests made with continuous loading and the "dashed" curve is for incremental
loading in which the rate of compression varied greatly during each increment.
The incremental loads have apparently caused structural damage and result in a
conservative estimate of preconsolidation. In practice this may just compensate
for the difference between laboratory and field rates of compression.

OVERCONSOLIDATION

The most heavily overconsolidated natural deposit of Leda clay reported

in the literature is from the lower St. Lawrence River valley. It has a measured
preconsolidation pressure of about 12 tons/sq, ft. (kg/cm2), a pressure consistent
with geological evidence of previous overburden (CONLON, 1966). Many other
parts of the valley contain similar soils which are more or less normally consolidated
under existing overburden.
It is often suggested that the measured preconsolidation pressure of these
soils may be due to a weathering action and not to overburden that has been
stripped off in geological time. In order to investigate this question, an assessment
was made of all the available test results on carefully chosen samples in a limited
area around Ottawa where the geological history is similar (CRAWFORD and EDEN,
1965).
Some of the samples were obtained in blocks from excavations and the
others were taken with special thin-walled piston samplers. Consolidation tests
were performed on specimens taken at elevations from 80 to 310 ft. above sea
level and the results are shown as dashed lines on Fig.7. Each line is a statistical
average of many tests on a single profile identified by a letter. Double letters
B B and K K indicate tests on block samples. Profile B was a deep excavation in
which it was possible to compare tests on block samples with tests on tube samples.
It may be noted that the overconsolidation measured on block samples is about
1 ton/sq, ft. (kg/cm 2) higher than that for tube samples.
Recent observations of piezometric levels have revealed a marked down-
ward flow of ground water in many areas. Seepage stresses may therefore account

Eng. Geol., 2(4) (1968) 239-265

250 C. B. CRAWFORD

I I I i I

3O0 ~ KEY __
A ~ A - --- Pc
E ~ ~ ~ ~ l i SU

I OKK - ~
"~ 200
g

B ~ ~B.~. ~C ~.
I00

I . t I 1 I
!,0 2,0 3.0 4.0 5,0
STRESS (KG/CM2)

Fig.7. Preconsolidation pressure Pc and undrained shear strength Su in relation to ele-

vation. Each line is a statistical average of many tests on a single profile identified by a letter;
double letters indicate tests on block samples (CRAWFORDand EDEN, 1965).

for part of the measured overconsolidation. Nevertheless, the observed relation-

ships between preconsolidation pressure and elevation are unquestionably related,
in general, to an inferred original sediment surface. The inferred surface is prob-
ably too high as it is above the present surface at all locations. This may be due
to superimposed seepage stresses, chemical action and to influences of compression
rate. F r o m a consideration of field and laboratory evidence the writer has con-
cluded that the fabric of the clay was established under low stresses and that weak
electrical links and chemical cementation between particles began immediately
to form a bonded structure. As the overburden increased, the structure was
compressed unidimensionally while still retaining its essentially card-house
orientation. Under these greater loads, the particles would be pushed closer to-
gether and additional contacts would be established. The chemical bonds between
particles in intimate contact would undoubtedly be much stronger than those
developed under small loads. This mechanistic picture appears to satisfy both the
external loading and chemical bonding concepts in relation to the response of the
present soil structure to stress changes.

OBSERVED SETTLEMENTS

Field observations of full-scale structures are essential for the reliable

assessment of the compressibility of natural soils, especially of highly compressible,
sensitive soils. For this reason one of the major projects of the Division of Building

Eng. GeoL, 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 251

Research of the National Research Council of Canada is the field measurement

of settlements.
The National Museum in Ottawa, built in 1910, is a classical settlement
failure. A differential settlement of nearly 2 ft. (60 cm) has left the structure
badly distorted. The building is a massive structure on spread footings with
bearing pressures ranging from 1 to 4 tons/sq, ft. (kg/cm2). Computed settlements
were generally much greater than the actual settlements (CRAwFORD, 1953).
This can be partly explained by the lack of satisfactory sampling and testing
equipment when the study was made about 15 years ago. A second factor in-
fluencing the accuracy of the settlement computation is the unknown degree of
load transfer within the structure.
Settlement observations on three apartment blocks in Ottawa were begun
in 1949. Each of the buildings rests on a reinforced concrete mat and the bearing
pressures range from 1.0 to 1.2 tons/sq, ft. (kg/cm2). Due to excavation the net
loading on the subgrade is only about 1/2 the bearing pressure applied by the mat
(i.e., averaging 0.6 ton/sq, ft.). The overconsolidation of the subsoil is about
1.6 tons/sq, ft. (kg/cm ~) and the observed settlement averaged about 0.5 inch
(1.3 cm) (LEGGET et al., 1961). Computed elastic settlement could account for
about 1/z of the measured settlement. Settlements computed from recompression
curves obtained in the consolidometer greatly overestimated the movements.
Large earth embankments provide ideal loading for research because the
increase in stresses in the subsoil can be computed with reasonable accuracy.
Three large embankments have been used for this purpose in the Ottawa area.
In two of the three, the applied stresses were in the recompression range.

Queensway, Green Creek

In 1957 an embankment 23 ft. high was placed on a thick deposit of moder-

ately overconsolidated Leda clay. The applied stress of more than 1 ton/sq, ft.
(kg/cm z) is equal to about half of the overconsolidation pressure. Settlement
gauges installed at various depths beneath the fill (BURN, 1959) showed that most
of the compression occurred in the upper 50 ft. The total settlement after 10 years
is approximately 5 inches (13 cm), half of which occurred immediately when the
load was applied (BURN and HAMILTON, 1968).

Queensway, Main Street

A second embankment, which also caused stress increases in the recom-

pression range, was built in 1963. Fill 22 ft. high was placed over 90 ft. of soil.
Four years after construction, the settlement at the base of the fill was about
4.5 inches (11 cm), of which nearly 3 inches (7 cm) occurred in the upper 30 ft.
of subsoil.

Eng. Geol., 2(4) (1968) 239-265

252 C. B. CRAWFORD

Kars Bridge Fill

The third embankment is located over only slightly overconsolidated clay.

The first 20 ft. of fill was placed in November 1959. In June 1961 the height was
increased to 26 ft. Settlement gauges placed at the base of the fill and at various
depths in the subsoil have provided a good record of the amount and rate of
compression of the natural ground. Piezometers under and adjacent to the loaded
area have revealed the build-up and dissipation of pore water pressures (EDEN
and POOROOSHASB,1968).
In April 1967 the base of the embankment had settled 20 inches (50 cm) at
the centre-line. The total thickness of soil is about 50 ft. but half of the compression
occurred in a soft clay layer extending from 20 to 35 ft. beneath the original
surface. When the first 20-ft. fill was completed (loading time of 15 days), the
piezometric excess pressure in the centre of this layer had increased to 27 ft. of
water. A month after the loading was completed the excess head had decreased
to 17 ft. of water. After 2 years the excess head was 11 ft., and during the past
4 years (to 1967) it has remained almost constant at 5 or 6 ft. It is of special interest
that very little pressure gradient has been observed between the centre of the
layer and its upper and lower boundaries. These observations lead to the conclusion
that the pore water pressures are being created by a breakdown of the soil structure
and that the rate of consolidation is therefore not dependent only on the perme-
ability of the soil.
In order to predict the settlement of this embankment two types of con-
solidation test were carried out. One set of samples was loaded by the normal
incremental procedure and a second set was loaded first to the overburden pressure,
then to the pressure of a 20-ft. fill, and finally to a pressure equivalent to a 26-ft.
fill. Both sets of tests indicated an ultimate settlement of about 3 ft. (90 cm) but
the latter type of test gave a predicted rate much closer to the actual rate. Obser-
vations to date indicate that the ultimate settlement will be almost 3 ft.
A record of settlement has also been obtained on a light single-storey building
resting on a gravel mat over about 40 ft. of almost normally consolidated clay at
Gloucester, near Ottawa. The building is located a few miles from the Kars
embankment described earlier. Although the total loading is small, it caused
stresses equal to the measured preconsolidation pressure in the subsoil and a
settlement of several inches has occurred.
A summary of settlements for the five case records is given in Table I. In
each case the tabulated values represent settlement after seven years. In cases 2, 3
and 5 settlements of individual layers of the subsoil were measured so that layers
of maximum compression during the period of observation could be selected.
In cases 1 and 4 the regions of maximum compression were estimated on the basis
of soil tests. These " m o s t complessible layers" were then treated as full scale
consolidation tests in order to assess the practical reliability of the preconsolidation

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 253

~ .-
~.~ o

- ~ 0

ud

.<
[..,

Eng. Geol., 2(4) (1968) 239-265

254 c.B. CRAWFORD

pressures measured in the laboratory. For this purpose the initial average effective
stress in the layer and the change in effective stress due to the applied loading were
calculated. The average preconsolidation pressure in the layer was estimated from
routine laboratory tests carried out usually in 5-7 days. The ratio of the final
effective stress to the average preconsolidation stress is used to illustrate relative
stress levels. Values up to 1 are in the recompression range and values greater than
1 are in the virgin compression range.
F o r convenience, average values for the three apartment buildings are
quoted because the subsoils were similar and the loading and settlements were
about the same in each case. The settlement measurements on the apartment
buildings were begun only after the slabs and basement columns were in place,
and the initial movements were, therefore, not observed.
The percentage settlement in the compressible layers is plotted with respect
to time after loading in Fig.8. The ratio of the average final effective stress to the
average preconsolidation stress is shown on each curve. It can be seen from these
curves that significant long term settlements occur at stresses well below the
measured preconsolidation pressure. This may be considered to be due to secondary
or delayed consolidation and the reasoning developed by BJERRUM (1967) appears
to be appropriate. The test results appear to be more realistic in practice when the
test is conducted slowly enough to incorporate an appreciable amount of secondary
consolidation.
An interesting record of the settlements measured at a large industrial plant
at Varrennes, Quebec has been published by CASAGm~NDE et al. (1965). The

~Apartments (0.56)
0
tr~
t~ r I
z

=
~2
_J

7~
.=,
a.

z4

m5
t~

I I I I I I "i
O

~6
1 2 3 4 5 6 7
TIME AFTER LOADING (YEARS)

Fig.8. T i m e - s e t t l e m e n t c u r v e s u n d e r field l o a d i n g .

Eng. Geol., 2(4) (1968) 2 3 9 - 2 6 5

QUICK CLAYS OF EASTERN CANADA 255

measured overconsolidation of the subsoil at the site was about 0.75 ton/sq, ft.
(kg/cm2). The applied load was sufficient to stress the subsoil up to the pre-
consolidation pressure. Measured settlements actually ranged from 1 to 6.5 inches
(2.5-16 cm). The authors attributed much of the settlement to virgin compression
caused by increased stresses due to lowering of the ground water table at the site.
In a discussion of the paper LEONARDS(1966) suggested that much of the measured
settlement might have been due to elastic type compression, consolidation in the
recompression range and secondary consolidation. It may be concluded that
computations of settlement for Leda clay are open to a variety of errors when the
maximum stresses are close to the measured preconsolidation pressure of the soil.

SHRINKAGE

Leda clay contains limited amounts of swelling clay minerals. At normal

water contents, however, its open structure will collapse and shrink substantially
when dried. The influence of clay concentration on shrinkage was studied by
mixing remoulded Leda clay with varying proportions of glass beads (DE JONG
and WARKENTIN, 1965). It was observed that both the amount of shrinkage and
the water content at the "shrinkage limit" decreased linearly with decreasing
percentage of clay as long as the glass beads were present as inclusions in the clay.
At clay concentrations of less than 30 ~ , the glass beads were in contact with one
another and very little shrinkage occurred.
The differential shrinkage of thin wafers of undisturbed Leda clay was
measured by WARKENT1N and BOZOZUK (1961). Total vertical and horizontal
shrinkage was found to be the same and this was attributed to initial random
particle orientation. The shrinkage limit decreased from 27 to 20 ~ when the soil
was remoulded, indicating a probable breakdown of the edge-to-face particle
arrangement.
Samples did not regain this original water content when rewetted but
the amount of regain increased as the amount of drying increased. The authors
concluded that shrinking increased the degree of parallel orientation and de-
creased the interparticle spacing, causing the diffuse layer of exchangeable ions
to overlap and to exert an osmotic swelling force between the approaching flat
surfaces. As the amount of drying increases, this osmotic swelling becomes more
prominent than the surface tension forces in the voids and is thought to control
the moisture regain.
Surface drying and wetting occurs seasonally and causes some volume
change in the natural clay. The greatest volume changes result from moisture
absorption by the roots of vegetation as they penetrate the high water content
clays at lower depths. Vertical ground movements were measured at various depths
and distances from a row of mature elm trees during a particularly hot and dry

Eng. Geol., 2(4) (1968) 239-265

256 c.B. CRAWFORD

summer at Ottawa in 1955 (BozozuK and BURN, 1960). Although the trees had
stood at the site for about 40 years, the combination of weather factors increased
their transpiration and decreased the available soil moisture to such an extent
that the ground surface near the trees settled 3 inches (8 cm) in 4 months. Even
at a depth of 6 ft. the vertical movement was 2 inches (5 cm). The influence of the
trees extended laterally for a distance almost equal to their height of 55 ft. (17 m).
Most of this settlement is not reversible. It was shown by these studies that vertical
ground movements can be related to "soil moisture depletion", a quantity that
can be estimated from weather records.
Shrinkage of the subsoil causes much damage to shallow foundations in
Leda clay. A survey of 574 structures in central Ottawa revealed the nature of
movements in relation to soil and vegetation conditions (BozozuK, 1962). Differ-
ential settlements up to 14 inches (35 cm) were observed in buildings 60-80 years
old. Little variation in soil properties was found but large movements were
always associated with the occurrence of large elms and other broad-leafed
trees. In most cases, the greatest settlement had occurred on the street side where
rows of trees are common and much of the precipitation is collected by the storm
sewers.
The distortion of a house resting on spread footings 5 ft. deep is shown
in Fig.9. At the left side of the photograph there is a line of mature trees and on the
right side is an open grassed area. Three curves in Fig.10 show the water content
profiles between the trees and the house, under the house, and under the grassed
area. The natural water content in the clay averages about 80 ~ . Within the region
of the trees' root system the water content has been reduced to less than 40
to a depth of 10 ft. (3 m). Under the house a sufficient reduction in water content
has occurred to cause the damage illustrated in the previous figure.
It has been concluded in general that if the upper 10 ft. of Leda clay has
a water content exceeding 40 ~ it is susceptible to shrinkage and special efforts
should be made to prevent differential drying (CRAWFORD, 1961a).

SHEARING RESISTANCE

The strength of Leda clay has attracted wide engineering interest because
of the unusual tendency of this material to liquefy and fail dramatically under
certain conditions. In its undisturbed state it varies from a very soft material with
undrained shear strengths as low as 0.15 ton/sq, ft. (kg/cm 2) (BuRkE and DAVIS,
1957) to a very hard material with shear strengths as great as 4 tons/sq, ft. (kg/cm~).
(CONLON, 1966). When disturbed it may have the consistency of a viscous liquid.
The undisturbed, undrained shear strength is, on the average, equal to about
3 0 ~ of the preconsolidation pressure (CRAWrORD and EDEN, 1965). The un-

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYSOF EASTERNCANADA 257

drained shear strength is compared in Fig.7 with the preconsolidation pressure at

several locations.
The undisturbed clay has a rigid structure. Good specimens will fail in
undrained compression at about 0.5 ~ strain. Elastic moduli as high as 1,500 tons/
sq. ft. (kg/cm e) have been measured in the laboratory (CoNLON, 1966). A modulus
of 720 tons/sq, ft. was computed from rebound measurements during an excavation
31 ft. deep and this value agreed with laboratory tests on specimens from the site
(BozozuK, 1963).
It is common practice to measure the undrained strength of Leda clay with
a field vane device. EDeN and HAMILXON (1957) found that strengths measured by
the field vane equaled or exceeded the maximum strengths obtained from laboratory

Fig.9. Building distortion due to shrinkage of the subsoil.

Eng. Geol., 2(4) (1968) 239-265

258 c.B. CRAWFORD

Depth of house f o u n d a t i o n s
5

6
i
7

.< 8

(/)
z

10

o 11

a_

13

>

c
a~
] I r I I I I
30 40 50 60 70 80 90
MOISTURE CONTENT (%)
Fig.lO. Variation in the moisture content of the subsoil.

compression tests. Vane test results were found to be more consistent than labo-
ratory results. Some evidence was presented to indicate that stability analyses based
on vane tests were satisfactory.
BAZETT et al. (1961) compared laboratory compression tests with field vane
tests in the study of a full-scale test trench failure. They found that the laboratory
tests agreed with the undrained failure analysis but use of the vane tests resulted
in a serious overestimate of the safety factor. The authors concluded that the vane
tests were accurate and attributed the discrepancy with field behaviour to loss of
strength due to lateral expansion of the material at the trench wall.
From a review of tests at nine sites near Ottawa, EDEN (1966) reaffirmed
that field vane tests are more consistent and yield higher strength values than
laboratory compression tests on soft to medium strength clays. For shear strengths
greater than 1 ton/sq, ft. (kg/cm 2) the field vane test was not considered to be
reliable. This conclusion was based on a comparison with compression tests on
specimens trimmed from block samples of strong, brittle clay. In this case it was
thought that the structure was partly destroyed by the insertion of the vane.
A series of experiments has been carried out in order to investigate the
influence of strain rate and stress path on the effective stresses and ultimate
shearing resistance of specimens of Leda clay (CRAWFORIg, 1959, 1961b, 1963a).
It has been shown that for overconsolidated specimens the maximum undrained

Eng Geol., 2(4) (1968) 239 265

QUICK CLAYS OF EASTERN CANADA 259

compressive strength decreases substantially within decreasing rate of strain. For

each 10-fold increase in the time to failure the ultimate strength decreases about
10 %. The slowest test recorded failed after 8 weeks at about 60 % of the strength
measured in a rapid test. Pore pressures with specimens under sustained constant
stress will increase due to breakdown of the sensitive soil structure, but if the
ambient stress is very low this does not appear to occur (COATES et al., 1963).
The lower strength measured in slow tests cannot therefore be attributed
to higher pore pressures, but must be due to a time-dependent reduction in true
cohesion. Measurement of pore water pressures at the ends and within soil speci-
mens during axial loading revealed higher pore pressures at the ends, and this
was attributed to disturbance caused by end restraint (CRAWFORD, 1963b).
For reliable pore pressure measurements it is necessary to use a back pressure
system and to conduct the test at less than 2 % axial strain/hour.
One set of unconfined compression tests was conducted on specimens
immersed in water in order to reduce the build-up of pore pressures during shear.
The average compressive strength at failure for three specimens was 2.4 tons/sq, ft.
(kg/cm2). They all failed at about 0.5 % strain after loading for more than 2 hours.
Almost all of the shearing resistance under these test conditions is due to cohesion
or interparticle bonds in the soil.
Research on the strength properties of Leda clay has been confined in
general to the evaluation and interpretation of friction and cohesion components
of the undisturbed soil. Specimens cut from block and tube samples have been sub-
jected to various stress paths and rates of strain while the response of the soil
skeleton to the effective stress changes was observed. A summary of the writer's
views (CRAWFORD, 1963a) led to the conclusion that most of the shearing re-
sistance of moderately overconsolidated Leda clay could be attributed to cohesion
caused by intrinsic stresses and interparticle bonds. When failure occurred at
stresses lower than preconsolidation pressure it was suggested that the use of the
Mohr-Coulomb theory might be inappropriate.
Recent work by KENNEY (1966) offers an alternative approach in which
shearing stress is related simply to the effective stress on the failure plane. This
value can be obtained from a variety of test methods and applied directly to a
practical problem. He has used this approach very successfully in analysing actual
landslides in quick clays in which cementation is a minor component. Kenney
defines friction as the strength component that is nearly directly proportional
to the applied effective stress and is mobilized through the mutual sliding of soil
particles irrespective of change in void ratio. By this definition the term may include
a substantial amount of cohesion according to the HVORSLEV (1960) concept.
Kenney also recognizes a second component due to cementation bonds, which
may be strain- and time-dependent and which he found to be of great significance
in Canadian clays from Labrador.
CONLON (1966) apparently used a similar concept when he divided the

Eng. Geol., 2(4) (1968) 239-265

260 ('. B. CRAWFORD

observed shearing resistance of clay from the lower St. Lawrence valley into two
components--cementation bond and friction--and arrived at a drained friction
angle of 36 °. This angle coupled with the cementation component appears to give
an unrealistic evaluation of the shear strength of the soil.

STABILITY PROBLEMS

Stability problems can be solved only by field research. Much research has
been done in recent years in an effort to correlate stress-deformation properties
of Leda clay with full-scale failure conditions. The bearing capacity of footings
on Leda clay is not in general a serious stability problem because excessive settle-
ments usually occur before shear failure. Settlements of several feet can occur
under buildings and earth embankments without bearing failure. There have been
some bearing failures of rapidly constructed fills, but they are not common.
The failure of an oil storage tank founded on soft, sensitive Leda clay has
been described by BROWN and PATERSON(1964). The tank, 44 ft. high and 70 ft.
in diameter, was placed on a 12-ft. sand mat over 35 ft. of clay. The tank sank
vertically 6-10 ft. within 24 hours of filling, although it had been partially filled
for several months. When full, the increase in vertical stress at the base of the sand
mat was estimated to be about 1.5 tons/sq, ft. (kg/cm2). The undrained shear
strength of the clay was about 0.25 ton/sq, ft. (kg/cm2), giving a computed factor
of safety of 1.2. The subsoil at the site was overconsolidated by only about 0.25
ton/sq, ft. (kg/cm z) and had an initial void ratio of about 2. It is evident, therefore,
that if the foundation had not failed due to shear it would have experienced several
feet of settlement.
The most serious stability problem is with natural slopes. Sometimes a
small bank failure will develop into an earth flow extending over a large area. In
some sections of the Ottawa-St. Lawrence valley the topography is dominated
by landslide scars such as shown in Fig.ll. A map showing major landslides in
the region has been published (HURTUBISE et al., 1957) and some of them have
been described (CRAW-FORD, 1961a).
Recent investigations of two flow slides near Ottawa have indicated that an
effective stress analysis using the method of slices for first failure gives a reasonable
approximation of the stability (CRAWFORD and EDEN, 1967). The influence of
field pore pressures has been found to be of the greatest importance in assessing
the stability. A similar investigation of a major landslide in the lower St. Lawrence
valley showed that it could not have been predicted by ordinary analyses (COYLON,
1966). In this case the clay was so strongly cemented that the landslide could be
explained only on the basis of a progressive failure initiated by dynamite blasting.
It was also noted that the landslide occurred in the springtime after an unusually
heavy rainy season.

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 261

Ground water is undoubtedly an important factor in these failures even if

they do not develop into flow slides. During the spring of 1967 after unusually
wet weather, there were many bank failures along the river banks at Ottawa.

Fig.11. Landslide scars in Leda clay (R.C.A.F. photograph).

Eng. Geol., 2(4) 0968) 239-265

262 c.B. CRAWFORD

Inspection soon after failure invariably revealed water running out of sandy or
silty layers in the fresh face of the slide.
A failure in a cut bank along the right-of-way for the Quebec North Shore
and Labrador railway was described by WOODS et al. (1959). The failure of the
70-ft. high bank occurred more than 2 years after it was cut on a 11/~-I slope. The
slide took place just 3 days after a 4-inch rainfall. A similar failure occurred in a
road cut near Ottawa in 1965. The cut, approximately 40 ft. deep, was made 5
years before the failure, which occurred after a night of heavy rains. A second failure
occurred 2 weeks later after another heavy rainstorm. After each failure great
quantities of water were released from the failure surface.
Analyses of these failures have not generally been satisfactory. Undrained
analyses usually result in a substantial overestimate of the safety factor. Effective
stress analyses are subject to error because of the unknown water conditions
and uncertainty of the failure mechanism. Even the case records referred to must
be treated with suspicion because the precise method of failure is unknown. Tension
cracks are not always taken into account and the initial failure surface does not
appear to agee with the computed critical surface.

CONCLUSIONS

The mineralogy and structure of the Canadian quick clays are similar to
those found in other parts of the world. Extensive deposits of the Leda clay have
a low salt content pore water and their sensitivity does not correlate with salt
content. Deposits which have been overconsolidated are strongly cemented.
Highly plastic deposits of Leda clay are very compressible. Although well-
defined pressure-void ratio curves can be obtained in laboratory tests the absolute
value of void ratio at any pressure is greatly dependent on the test procedure. In
normal tests (lasting 5 days) a considerable amount of secondary consolidation
is permitted. Field observations show that much secondary consolidation occurs
in nature. The available evidence suggests that a 5-day test will overestimate the
preconsolidation pressure of a good sample but that this may be compensated
by disturbance of tube samples. A significant amount of compression has been
observed in the field at pressures well below the preconsolidation pressure in the
ground. Field observations indicate that pore pressures are created by breakdown
of the soil structure as it compresses under load and this may compensate for
pressure decreases by drainage.
Large settlements occur due to shrinkage in the upper 10 or 15 ft. of Leda
clay. The shrinkage is caused primarily by the withdrawal of soil water by vege-
tation, especially by trees. Shrinkage potential is related to the natural water con-
tent. The problem can be avoided by restricting tree growth near shallow foun-
dations. Unless the soil is dried to a very low water content it has little swelling
potential.

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 263

T h e strength of Leda clay varies from very soft to very stiff. The measured
u n d r a i n e d strength depends o n the rate of testing. M o s t of the shearing resistance
u n d e r low stresses has been attributed to cohesion. The most difficult strength
problems are e n c o u n t e r e d in a t t e m p t i n g to assess the stability of slopes in over-
consolidated deposits. M u c h field research will be required to solve this p r o b l e m
satisfactorily.

ACKNOWLEDGEMENTS

This paper is a review of the research of m a n y workers. The a u t h o r is

especially grateful for the assistance of his colleagues in the Division of Building
Research of the N a t i o n a l Research C o u n c i l of C a n a d a a n d for the e n c o u r a g e m e n t
of the Director, Dr. R. F. Legget, with whose a p p r o v a l it is published.

REFERENCES

ALLEN, V. T. and JorrNs, W. D., 1960. Clays and clay minerals of New England and eastern
Canada. Bull. Geol. Soc. Am., 71: 75-86.
BAZETT, O. J., ADAMS,J. I., and MATYAS,E. L., 1961. An investigation of slides in a test trench
excavated in fissured sensitive marine clay. Proc. Intern. Conf. Soil Mech. Found. Eng.,
5th, Paris, 1960, 1: 431-435.
BJERRUM, L., 1954. Geotechnical properties of Norwegian marine clays. Geotechnique, 4: 49-70.
BJERRUM, L., 1967. Engineering geology of Norwegian normally-consolidated marine clays as
related to settlements of buildings. Geotechnique, 17: 83-117.
BozozuK, M., 1962. Soil shrinkage damages shallow foundations at Ottawa, Canada. Eng.
J., 45: 33-37.
BOZOZUK, M., 1963. The modulus of elasticity of Leda clay from field measurements. Can.
Geotech. J., 1: 43-51.
BozozuK, M. and BURN, K. N., 1960. Vertical ground movements near elm trees. Geotechnique,
10: 19-32.
BROWN, J. O. and PATERSON,W. G., 1964. Failure of an oil storage tank founded on a sensitive
marine clay. Can. Geotech. J., 1: 205-214.
BRYDON, J. E. and PATRY, L. M., 1961. Mineralogy of Champlain Sea sediments and a Rideau
clay soil profile. Can. J. Soil Sci., 41: 169-181.
BURKE, H. and DAVIS,W., 1957. Physical properties of marine clay and their effect on the Grass
River Lock excavation. Proc. Intern. Conf. Soil Mech. Found. Eng., 4th, London, 1956,
2: 301-304.
BURN, K. N., 1959. Instrumentation for a consolidation study of a clay deposit beneath an
embankment. Geotechnique, 9: 136-142.
BURN, K. N., 1966. The effect of iron in the determination of moisture content by the neutron
method. Can. J. Earth Sci., 3: 129.
BURN, K. N. and HAMILTON,J. J., 1968. Settlements of a high embankment on Ledaclay.--Can.
Soil Mech. Conf., 20th, Quebec, 1967--Can. Geotech. J., 5: 16-27.
CASAGRANDE,L., FIRING,P., ScI-IOOr, G., and TURCKE,E. W. J., 1965. Settlement of mat foun-
dation on thick stratum of sensitive clay. Can. Geotech. J., 2: 299-312.
COATES, D. F., BURN, K. N., and McRosTtE, G. C., 1963. Strain-time-strength relationships
in a marine clay. Trans. Eng. Inst. Can., 6" 12pp.
CONLON, R. J., 1966. Landslide on the Toulnustouc River, Quebec. Can. Geotech. J., 3: 113-144.

Eng. Geol., 2(4) (1968) 239-265

264 c . B . CRAWFORD

CRAWrORD, C. B., 1953. Settlement studies on the National Museum Building, Ottawa, Canada.
Proc. Intern. Conf. Soil Mech. Found. Eng., 3rd, Zurich, 1952, I : 338-345.
CRAWFORD, C. B., 1959. The influence of rate of strain on effective stresses in sensitive clay.
Ant. Soc. Testing Mater., Spec. Tech. Publ., 254: 36-61.
CRAWFORD, C. B., 1961a. Engineering studies of Leda clay. In: R. F. LEGGET (Editor), Soils in
Canada. Roy. Soc. Can., Spec. Publ., 3: 20~217.
CRAWFORD, C. B., 1961 b. The influence of strain on shearing resistance of sensitive clay. Proc.
Am. Soc. Testing Mater., 61:1250 1276.
CRAWEORD, C. B., 1963a. Cohesion in an undisturbed sensitive clay. Geotechnique, 13:132 146.
CRAWFORD, C. B., 1963b. Pore pressures within soil specimens in triaxial compression. Am. Soc.
Testing Mater., Spec. Tech. Publ., 361:192-21 I.
CRAWEORD, C. B., 1964. Interpretation of the consolidation test. J. Soil Mech. Found. DIE.,
Proc. Am. Soe. Civil Engrs., 90: 87-102.
CRAWFORD, C. B., 1965. Resistance of soil structure to consolidation. Can. Geoteeh. J., 2:90-115.
CRAWFORD, C. B. and EDEN, W. J., 1965. A comparison of laboratory results with in situ properties
of Leda clay. Proc. Intern. Conf. Soil Mech. Found. Eng., 6th, Montreal, 1964, 1: 31-35.
CRAWFORD, C. B. and EDEN, W. J., 1967. Stability of natural slopes in sensitive clay. J. Soil
Mech. Found. DIE., Am. Soc. Civil Engrs., 93: 419-436.
DE JONG, E. and WARKENTIN, B. P., 1965. Shrinkage of soil samples with varying clay con-
centrations. Can. Geotech. J., 2: 16-22.
EDEN, W. J., 1966. An evaluation of the field vane test in sensitive clay. Am. Soc. Testing Mater.,
Spec. Tech. Publ., 399: 8-17.
EDEN, W. J. and CRAWFORD, C. B., 1957. Geotechnical properties of Leda clay in the Ottawa
area. Proc. Intern. Conf. Soil Mech. Found. Eng., 4th, London, 1957, 1: 22-27.
EDEN, W. J. and HAMILTON, J. J., 1957. The use of a field vane apparatus in sensitive clay.
Symp. In-place Shear Testing Soil Vane Method--Am. Soe. Testing Mater., Spee. Tech.
Publ., 193: 41-53.
EDEN, W. J. and KUBOTA, J. K., 1961. Some observations on the measurement of sensitivity of
clays. Proe. Am. Soc. Testing Mater., 61: 1239-1249.
EDEN, W. J. and POOROOSHASB, H. B., 1968. Settlement observations on Kars bridge approach
fill. Can. Soil Mech. Conf., 20th, Quebec City, 1967--Can. Geoteeh. J., 5: 28~15.
GADD, N. R., 1962. Surficial geology of Ottawa map-area Ontario and Quebec. Geol. SurE.
Can., Paper, 62(16): 4 pp.
GILLOTT, J. E., 1968. Clay in Engineering Geology. Elsevier, Amsterdam, in press.
HAMILTON, J. J. and CRAWFORD, C. B., 1959. Improved determination of preconsolidation pres-
sure of a sensitive clay. Am. Soc. Testing Mater., Spec. Tech. Publ., 254: 254-271.
HURTUBISE, J. E., GADO, N. R. and MEVERHOF, G. G., 1957. Les 6boulements de terrain dans
l'est du Canada (Landslides in eastern Canada). Proe. Intern. Conf. Soil Mech. Found.
Eng., 4th, London, 1957, 2: 325-329.
HVORSLEV, M. J., 1960. Physical components of the shear strength of saturated clays.--Research
conf. shear strength of cohesive soils, Boulder, Colo. Proc. Am. Soc. Civil. Engrs., 169-273.
JARRETT, P. M., 1967. Time-dependent consolidation of a sensitive clay. Mater. Res. Std., Am.
Soc. Testing Mater., 7: 300-304.
JOHNSTON, W. A., 1917. Pleistocene and recent deposits in the vicinity of Ottawa, with a de-
scription of the soils. Geol. SurE. Can., Mere., 101 : 69 pp.
KARROW, P. F., 1961. The Champlain Sea and its sediments. In: R. F. LEGGET (Editor), Soils
in Canada. Roy. Sac. Can., Spec. Publ., 3: 97-108.
KENNEY, T. C., 1964. Sea-level movements and the geologic histories of the post-Glacial marine
soils at Boston, Nicolet, Ottawa and Oslo. Geotechnique, 14: 203-230.
KENNEY, T. C., 1966. Shearing Resistance of Natural Quick Clays. Thesis Univ. London, London,
138pp.
LEGGET, R. F., BURN, K. N., and BOZOZUK, M., 1961. Three buildings on "floating foundations"
in Ottawa, Canada. J. Natl. Buildings Organ., 6: 106-119.
LEONARDS, G. A., 1966. Discussion of "settlement of mat foundation on thick stratum of sensitive
clay". Can. Geotech. J., 3: 100-102.
PENNER, E., 1963a. Sensitivity in Leda clay. Nature, 197: 347-348.

Eng. Geol., 2(4) (1968) 239-265

QUICK CLAYS OF EASTERN CANADA 265

PENNER, E., 1963b. Anisotropic thermal conduction in clay sediments. In: I. T. ROSENQVISTand
P. GRAFF-PETERSEN(Editors), Proc. Intern. Clay Conf., 1963. Pergamon, London, 1:
365-376.
PENNER, E., 1964. Studies of sensitivity and electrokinetic potential in Leda clay. Nature, 204:
808-809.
PENNER, E., 1965. A study of sensitivity in Leda clay. Can. J. Earth Sci., 2: 425-441.
QUIGLEY, R. i . and THOMPSON, C. D., 1966. The fabric of anisotropically consolidated sensitive
marine clay. Can. Geotech. J., 3: 61-73.
ROSENQVIST, I. T., 1955. Investigations in the clay-electrolyte-water system. Norweg. Geotech.
Inst. Rept., 9: 125pp.
WARKENTIN,B. e. and BOZOZUK, M., 1961. Shrinking and swelling properties of two Canadian
clays. Proc. Intern. Conf. Soil Mech. Found. Eng., 5th, Paris, 1960, 1: 851-855.
WOODS, K. B., PRYER, R. W. J., and EDEN, W. J., 1959. Soil engineering problems on the Quebec
north shore and Labrador railway. Am. Ry. Eng. Assoc., Bull., 60: 669-688.

Eng. Geol., 2(4) (1968) 239-265