Briefing: Skempton's legacy to the Proceedings of the Institution of Civil Engineers

Professor Alec Skempton (1914–2001) wrote a total of nine papers for the Proceedings of the Institution of Civil Engineers (and its predecessors), during the period 1941–1962. His coauthors and correspondents have included luminaries such as Cooling, Terzaghi, Golder, Meyerhof, Fellenius, Glossop, Macdonald and Peck. Amongst the subjects that have caught his attention are the behaviour of clay, bearing capacity and settlement analysis of buildings, and clay shrinkage caused by trees.

Skempton was involved in 17 discussions and wrote correspondence on seven occasions. The majority of his work appeared in Géotechnique, following its establishment in 1948.

A full list of Skempton's writings in the Proceedings is given below and his very first paper, written with Cooling in 1941, is reproduced for historical interest at the end.

Cooling L. F. and Skempton A. W. (1941)

Some experiments on the consolidation of clay

Journal of the ICE, 16, No. 7, 1940-41, June 1941, 381–398.

Cooling L. F. and Skempton A. W. (1942)

A laboratory study of London clay

Journal of the ICE, 17, No. 3, 1941–42, Jan. 1942, 251–276.

Skempton A. W. (1942)

An investigation of the bearing capacity of a soft clay soil

Journal of the ICE, 18, No. 7, 1941–42, June 1942, 307–321.

Skempton A. W. (1945)

A slip in the West Bank of the Eau Brink cut

Journal of the ICE, 24, No. 7, May 1945, 267–287.

Glossop R. and Skempton A. W. (1945)

Particle-size in silts and sands

Journal of the ICE, 25, No. 2, Dec. 1945, 81–105.

Skempton A. W. (1954)

A foundation failure due to clay shrinkage caused by poplar trees

ICE Proceedings, General, Part I, 3, Jan. 1954, 66–86.

Skempton A. W., Peck R. B. and Macdonald D. H. (1955)

Settlement analyses of six structures in Chicago and London

ICE Proceedings, General, Part I, 4, July 1955, 525–542.

Skempton A. W. (1955)

Foundations for high buildings

ICE Proceedings, Eng. Divisions (HPSW), Part III, 4, Aug. 1955, 246–269.

Skempton A. W. and Macdonald D. H. (1956)

The allowable settlements of buildings

ICE Proceedings, Eng. Divisions (HPSW), Part III, 5, Dec. 1956, 727–768.

Cooling L. F., Markwick A. H. D., Clements R. G. H., Lyddon A. J., Pimm G. B. R., Jackson J. S., Lowe Brown W. L, Wilson G., Harding H. J. B., Allin R. V., Bennett G. T., Bailey E. B., Du Plat Taylor F. M., Glossop R., Golder H. Q. and Skempton A. W. (1942)

Discussion on soil mechanics and site exploration and on soil mechanics in road and aerodromes construction

Journal of the ICE, 18, No. 6, 1941–42, April 1942, 155–180.

Anderson D., Binnie W. J. E., Brand C. G., Halcrow W. T., Robertson V. A. M., Harding H. J. B., Skempton A. W., Cox S. W., Segrave J. H., Gardner C. T. and Boycott G. W. M. (1943)

Tunnel linings, with special reference to a new form of reinforced concrete lining

Journal of the ICE, 20, No. 5, 1942–43, March 1943, 42–64.

Wentworth Sheilds, Knight C. W., Du Plat Taylor F. M. G., Wilson J. S., Cooling L. F., Packshaw S., Skempton A. W., Manning G. P., Bickley J., Palmer J. E. G. Hiley (1944)

The design of wharves on soft ground

Journal of the ICE, 22, No. 5, 1943–44, March 1944, 25–45.

Harding H. J. B., McHaffie M. G. J., Guthrie Brown 1, Brown J. G., Prior F. E., Cooling L. F., Cassel F. L, Toms A. H., Rooertshaw J., Brown C. B., Murdock L. J., Hammond R., Le Grand J. P., Rodin S., and Skempton A. W. (1949)

Site investigations including boring and other methods of sub-surface exploration. Works Construction. Paper no 12.

Journal of the ICE, 32, No. 6, April 1949, 137–157.

Wilson G., Morgan H. D., Skempton A. W., Bickley J., Marshal C. C, Meyerhof G. G., Scott P. A., Little D. H., Boulton N. S., Wood G. (1950)

The bearing capacity of screw piles and screwcrete cylinders

Journal of the ICE, 34, No. 5, March 1950, 74–93.

Hartley A. C, Golder H. Q., Skempton A. W., Murdock L. J., McHardy Young J., Little D. H., O'Sullivan T. P., Barnes E. M., Saurin B. F. and Pike C. W. (1952)

Buoyant foundations in soft clay for oil-refinery structures at Grangemouth

ICE Proceedings, Eng. Divisions (PHSW), Part III, 1, Dec. 1952, 321–334.

Skempton A. W., Gardner G. A., Bondy O., Davey N., Mears R. P., Golder H. Q., Capper P. L. and Wilkins R. 1, (1952)

The historical development of structural theory

ICE Proceedings, Eng. Divisions (PHSW), Part III, 1, Dec. 1952, 402–419.

Bradbeer B. F. J., Maclean D. J., Golder H. Q., Toms A. H., Morgan E. E., Cassel F. L., Skempton A. W., Fairbairn P. E., Craven C. A. U., Fox D. A., Jakobson B., Ward W. H. and Bryan B. W. (1953)

Soil stability problems in road engineering

ICE Proceedings, Eng. Divisions (AMRR), Part II, 2, June 1953, 254–280.

Cousins H. G., Bowie P. G., Brown C. B., Measor E. O., Morice P. B., Powers M. A. R., Tottenham H., Leonard M. W., Littel D. H., Skempton A. W., Cockle W. P. S., Baker A. A. L, Amp O. and Jenkins R. S. (1953)

The design of a reinforced-concrete factory at Brynmaur, South Wales

ICE Proceedings, Eng. Divisions (HPHSW), Part III, 2, Dec. 1953, 380–397.

Clayton A. J. H., Dracker D. G., Denby E., Jensen R. R. A., Cooling L. F., Skempton A. W., Sanyal N., O'Sullivan T. P., Golder H. Q., Stamp D., Visvevaraya H. C, Wood R. H., Meyerhof G. G., Measor E. O., Home M. R. and Derrington J. A. (1955)

High buildings-the traffic and parking problems

ICE Proceedings, Eng. Divisions (HPSW), Part III, 4, Aug. 1955, 275–313.

Malcolm J. R., Allen F. H., Wild G. E., Sivewright W. J., Reed A. L., Dempsey J. A., Peel C, Skempton A. W., Cornfield G. M. and Burns T. F. (1956)

Reconstruction of the Gallions lower entrance lock at the Royal Docks of the Port of London Authority

ICE Proceedings, Eng. Divisions (AMRR), Part II, 5, Feb. 1956, 156–169.

Measor E. O., Cooling L. F., Souza R. W., Golder H. Q., Williams G. M. J., Ward W. H., Cassel F. L, Flint A. R., Meyerhof G. G., Terzaghi K., Ripley C. F., Skempton A. W., Peck R. B., Deere D. U., Capacete J. L, Schriever W. R. and Plewe (1956)

The allowable settlements of buildings

ICE Proceedings, Eng. Divisions (HPSW), Part III, 5, Dec. 1956, 768–784.

Legget R. F., Skempton A. W., Cooling L. F., Kellaway G. A., Little A. L, Lea N. D., Macdonald D. A. E., Ward W. H., Sutherland H. B., Brown J. B., Cassel F. L., Mak W. K., Hardy D. R. M., Knibb J. G., Boswell P. G. H. and Lawton F. L. (1959)

Soil engineering at Steep Rock Iron Mines, Ontario, Cananda

ICE Proceedings, 13, May 1959, 93–117.

Pippard A. J. S., Cassie W. F., Morgan H. D., Brand C. G., Newmark N. M., Dillon E. C, McDonald A, Sparkes S. R., Hendry A. W., Brind H. G., Owen J. B. B., Evans R. H., Marshal W. T., Broadbent B. H.,. Skempton A. W. and Vaughan, S. (1959)

The place of the university in the education of civil engineers

ICE Proceedings, 12, April 1959, 7–16.

Campbell H. E., Skempton A. W., Newmark N. M., Amp O. N., Mostertman L. J., Sparkes S. R., Owen J. B. B., Roberton J. F. L., Evans R. H., Gwynn J. D., Morice P. B., Boulton N. S., Andrew R. P. and Naylor A. H. (1959)

The place of the university in the education of civil engineers

ICE Proceedings, 12, April 1959, 16–23.

Morris S. S., Law R. U., Herbert W. J., Cronin H. F., Little A. L, Myburgh R. I., Kleynhans J. H., Rangeley W. R., Fleming J. H., Banks J. A., Kennard M. F., Skempton A. W., Martin J., Humphreys J. D., Shilston A. W. and Stallebra (1960)

The Cape Town Wemmer-Shoek water scheme

ICE Proceedings, 15, Feb. 1960, 146–165.

Ischy E., Glossop R., Skempton A. W., Leonard M. W., Walters R. C. S., Kell J., Perrott W. E., Skipp B. O., Greenwood D. A., Little A. L, Morgan H. D., Lamberton B. A. and KennardM. F. (1962)

An introduction to alluvial grouting

ICE Proceedings, 23, Dec. 1962, 705–725.

Hunter J. W. and Skempton A. W. (1942)

The ultimate bearing pressure of rectangular footings

Journal of the ICE, 18, Oct. 1942, Supplement to No. 8, 1941–42, 458–463.

Cooling L. F., Skempton A. W., Markwick A. H. D. and Walker E.G. (1942)

A laboratory study of London clay

Journal of the ICE, 18, Oct. 1942, Supplement to No. 8, 1941–42, 497–501.

Skempton A. W., Doran W. E., Lowe Brown W. L., Terzaghi K., Ward W. H., Wilkinson E. and Wilson G. (1945)

A slip in the west bank of the Eau Brink cut

Journal of the ICE, 24, Supplement to No. 8, Oct. 1945, 535–553.

Glossop R., Skempton A. W., Boswell P. G. H., Campus F., Clare K. E., White E. E., Edwards R. M. W., Daxelhofer J. P., Golder H. Q., Huizinga T. K., King J. W. H., Lacey G., Wilson G., Murdock L. J., Packshaw S. and Terzaghi K. (1946)

Particle-size in silts and sands

Journal of the ICE, 26, Supplement to No. 8, Oct. 1946, 557–580.

Faber O., Adcock A. R. W., Allin R. V., Blake F. H., Boycott G. W. M., Skempton A. W., Worth R. E. J., Edwards R. M. W., Penney W. G., Cassel F. L, Golder H. Q., Hunter J. W., Owen Lake J., Turner L., Wynne Edwards R. M. and Ledson (1947)

A new piling formula

Journal of the ICE, 28, Supplement to No. 8, Oct. 1947, 522–570.

Skempton A. W., Elgar W. H., Harris D. A., King J. H. G., Creswell D. A. and Legget R. F. (1954)

A foundation failure due to clay shrinkage caused by poplar trees

ICE Proceedings, General, Part I, 3, Sept. 1954, 604–621.

Skempton A. W., Peck R. B., Macdonald D. H., Hayard E. M., Alderman J. K. and Meyerhof G. G. (1956)

Settlement analyses of six structures in Chicago and London

ICE Proceedings, General, Part I, 5, March 1956, 166–172.

5. SKEMPTON'S FIRST PAPER IN ICE PROCEEDINGS^*

The consolidation of clays was first studied by Professor Karl von Terzaghi, M. Inst. C.E.¹, and his theory of consolidation has proved of great value in problems of foundation design involving settlement².

The term “ consolidation ”, as used in soil mechanics, refers to the process by which a soil undergoes a decrease in volume when it is subjected to an increase in pressure. This volume-reduction, or compression, is brought about by a closer packing of the grains, and its magnitude depends on the pressure-increment and on the compressibility of the soil. With fine-grained soils such as clays and silts, the voids of which are completely filled with water, the decrease in volume follows only gradually on the application of pressure, and a considerable time elapses before the soil reaches a state of equilibrium.

It was to explain this time-lag that Terzaghi advanced his theory of consolidation. He pointed out that as the pores are completely filled with water, a volume-reduction can occur only by the escape of a corresponding volume of water from the clay into a free drainage surface. He suggested that the mechanism of the process was as follows. The pressure-increment is first taken wholly by the pore water, thus setting up in the pore water an excess hydrostatic pressure. At a drainage surface the pore water has zero excess pressure, and consequently a hydraulic gradient is set up, which in turn causes water-movements. The consolidation process is then governed by the gradual dissipation of the excess hydrostatic pressure and, as fine-grained soils offer a high resistance to the flow of water, the process will be slow. When the excess pressure has finally become equal to zero the soil is in equilibrium under the applied pressure. This pressure is then said to be “ effective.”

In the mathematical development of his theory, Terzaghi made the simplifying assumption that the flow of water is one-dimensional only, a condition which would obtain for a layer of clay lying between two sand layers, or between one sand layer and a bed of shale or rock. He was then able to show that the consolidation process could be represented by a partial differential equation of the second order, similar in form to the Fourier law of linear diffusion, and involving a constant for the material termed the “ coefficient of consolidation.” The solution of this equation requires that the position of the drainage surfaces and the pressure-distribution in the layer should be known. For a full treatment of this analysis reference should be made to the standard work by Terzaghi and Fröhlich¹. In the present Paper only the results of the analysis will be outlined.

Terzaghi also carried out a series of tests to study the extent to which the theory represents the actual consolidation process in clays. For this purpose he developed the “ oedometer ”, an apparatus in which a clay layer was subjected to a load under conditions similar to those postulated in the theory, that is, resulting in a one-dimensional flow of water. For the clays with which he was working, he was able to show that the experimental time/consolidation curves were of the same general shape as the theoretical curves, except that they showed some deviation in the later stages of consolidation². He also stated, however, that with some soils the theory was not closely followed.

Since that time many oedometer tests have been made, but very few accounts of comparisons between experimental and theoretical results have been published. The general conclusion of the Harvard laboratory³, in 1936, was that the deviations in the later stages of consolidation occur to a greater or less degree in all soils, but are most pronounced in soils containing organic matter.

In this Paper a brief outline is given of the oedometer test, with descriptions of a few simple experiments on samples prepared from a stock of London clay made homogeneous by remoulding with water. The tests serve to show that London clay yields results in close agreement with theory, and also demonstrate in a straightforward manner some of the main conclusions of the theory.

A modern form of oedometer, developed by Terzaghi and Casagrande, is illustrated diagrammatically in Fig. 1 . The sample of clay is held in a brass ring between porous stones, which provide a free drainage surface for the water expelled from the sample during consolidation. The stones consist of disks about ½ inch thick, of an open-pored Portland limestone, and they are kept in contact with water to prevent the sample from drying out during the test. The samples used in a standard test are 3 inches in diameter and ¾ inch thick, but oedometers of 4-inch diameter, taking samples of variable thickness, are also in use. Pressure is applied to the sample through the upper porous stone by placing a weight on the hanger at the end of a lever-arm. The lever system is counterbalanced so that very small pressures can be used, and is arranged so that a weight of 10 lb. on the hanger applies a pressure of 1 ton per square foot on the sample.

Since the sample is restrained laterally, the consolidation process can be followed by observing its decrease in thickness. This is done by means of a dial gauge reading to 1 × 10⁻⁴ inch. On applying a pressure, compression takes place, rapidly at first, but the rate decreases gradually with time, as shown in Fig. 2 .

After an interval, largely depending upon the thickness of the sample, the dial movements practically cease, and the sample can then be considered to be in equilibrium.

The final compression of the sample under this pressure is denoted by s_∞ (since theoretically an infinite time is required for equilibrium to be reached) and the compression at any time t after application of the pressure is denoted by s_t. The degree of consolidation μ at this time is defined by the equation

and the s/t curve of Fig. 2  can readily be plotted as a μ/t curve, as in Fig. 3 .

After an approximately steady state has been reached the pressure is increased, when a similar consolidation process occurs and a new equilibrium-density of the clay under the increased pressure is attained. Successive pressure-increments are applied, and give a corresponding number of equilibrium-densities. The density of a sample is expressed as the voids ratio, defined by the equation

and the relation between effective pressure and voids-ratio (the ϵ/p curve) is shown in Fig. 4 .

The oedometer test therefore gives one pressure/voids-ratio curve and a number of time/consolidation curves.

The p/ϵ curve is a characteristic of the clay, and from it the compressibility can be determined. The compressibility, for linear consolidation, is defined as the decrease in thickness per unit thickness per unit pressure-increase, and can be determined in the following manner.

A clay sample has an original thickness l₁ and a voids ratio ϵ₁, when in equilibrium under a pressure p₁. If the pressure be increased by σ the sample consolidates to a final thickness l₂ and voids ratio ϵ₂ under the pressure p₂ = p₁ + σ. Now from equation (1),

but in the test the decrease in volume equals the decrease in volume of voids: hence

and as the area of the sample remains constant,

The compressibility is, therefore, given by the equation

Denoting the average slope of the p/ϵ curve over the increment σ by a,

The compressibility is not, in general, a constant, but is a function of ϵ and hence of p. From the p/ϵ curve the value of $α1+ϵ$ can, therefore, be calculated for any value of ϵ or p, and the p/ϵ curve also enables the final compression of a clay layer to be calculated for any pressure increment, from equation (2).

The rate at which the consolidation proceeds is, however, a more complicated problem, and it is to this problem that the theory of consolidation is applied. The principal result of this theory is that the degree of consolidation μ is a function only of the factor (ct/d²).

In this factor c is the “ coefficient of consolidation ”, a constant for a given clay, t denotes the time after application of the load causing consolidation, and d denotes the “ drainage path ”, that is, the maximum distance which water has to travel in the clay (measured in a straight line) before reaching a free drainage-surface. In the oedometer test d is one-half the thickness of the sample, as the water can drain from both faces into the porous stones.

The above result may be written in the form

where τ denotes the “ time factor ”, which is dimensionless, and c has the dimensions L². T⁻¹.

The function ψ depends upon the distribution of pressure in the layer and upon whether one or both faces of the layer constitute drainage-surfaces. For the particular conditions of the oedometer test, namely, drainage from both faces of the layer and a uniform distribution of pressure, the solution is

Values of μ calculated from the above equation for some particular values of τ are given in the following Table :—

Now it follows from equation (3) that, for any given conditions of pressure and drainage, and for a particular clay, the value of μ at any given time varies inversely as d². In order to confirm this result in a simple manner, oedometer tests were carried out on three samples of the same clay with different thicknesses.

A further check on the degree of validity of the theory is rendered possible by comparing the shape of the experimental time/consolidation curves with the curve defined by equation (4). This is the comparison which was referred to in the introduction. As will be seen later, the comparison is made a relatively simple matter by using the result that in the range 0 <μ <0·5, μ is proportional to $τ.$ This can be found from equation (4), but it has been shown more directly by Dr. E. N. Fox, of the Building Kesearch Station, who has derived the equation

This equation is true to a high degree of accuracy within the range indicated, and was obtained by the method of the operational calculus, as applied by Dr. H. Jeffreys¹ to problems in the conduction of heat, which are analogous mathematically to those in the consolidation of clays.

In order to carry out the experiments with material as homogeneous as possible, about 1 cubic foot of London clay was remoulded very thoroughly with water to form a stiff slurry and was allowed to stand for several weeks under damp cloths in order further to equalize the water-content distribution. From this mass of clay small samples were taken, and on these samples various simple tests, principally for classification purposes, were carried out. For the methods of carrying out these tests, and for their physical significance, reference should be made to a recent text-book² on soil mechanics. Very little variation was found to exist between the different samples, and the mean results were as follows :—

The above values indicate a typical inorganic clay.

The clay slurry was placed in three oedometers, one of 3-inch diameter and two of 4-inch diameter, until the thickness of clay was about 1·0 inch, 1·6 inch, and 3·1 inches respectively. These samples will be referred to as Samples I, II, and III. The clay was gently tamped into position, to exclude as far as possible any small air-bubbles, although it was not to be expected that they could be avoided entirely.

The upper porous stone was then placed on the surface of the clay, which had been carefully smoothed off, and a pressure of about 0·1 ton per square foot was applied through the piston. The dial-gauge showed that the clay at once started consolidating, although part of the movement under this load was due to a slight extrusion of slurry into the space between the stone and the brass ring. After the sample had been subjected to this pressure for 1 day, the pressure was increased to 0·5 ton per square foot, which was maintained until the dials showed that the rate of movement had become sensibly zero. With Sample I this occurred after 4 days, whilst 10 days were allowed for Samples II and III, the movements during the last 24 hours being 3 × 10⁻⁴ inch, 3 × 10⁻⁴ inch, and 8 × 10⁻⁴ inch respectively. Under this pressure the clay became quite firm.

The pressure was then successively increased on each sample from 0·5 ton per square foot to 1 ton, 2 tons, and 4 tons per square foot, each increment being maintained for 3, 5, and 9 days on Samples I, II, and III. Dial readings were taken throughout: the final compressions under each increment are given in Table I. The actual observations for the three samples under the increment from 1 ton per square loot to 2 tons per square foot are plotted in Fig. 5 .

The average compressions during the last 24 hours of each increment were 2 × 10⁻⁴ inch, 5 × 10⁻⁴ inch, and 1 × 10⁻⁴ inch for the three samples, or less than 1 per cent, of the final compression movements.

When each sample had come to this approximate equilibrium under the greatest pressure of 4 tons per square foot, the load was removed, the oedometer was dismantled, and the sample was weighed (W_f); it was then placed in a ventilated oven maintained at 105° C. and periodically reweighed, after being allowed to cool in a desiccator. When a constant weight W₀ had been reached, it was assumed that all the water had been driven off, and the water-content of the sample when in equilibrium under 4 tons per square foot was then determined from the relation

The results of 33·7, 33·8, and 33·4 for samples I, II, and III, are probably fortuitously close, although a reasonable agreement would be expected on account of the homogeneity of the material.

From the definition of voids-ratio given in equation (1) it follows that, if the voids are saturated with water, the value of ϵ under 4 tons per square foot is given by

The three values are plotted in Fig. 6 . (The slight differences in pressure are due to the pressure on all three samples not being exactly equal, but the differences are negligibly small.)

The thickness of the samples l_f under this pressure can also be calculated, for if A denotes the area of the sample and γ_w the density of water, then

These values are given in Table I, and from the observed final compressions it is at once possible to calculate the thickness under the other pressures. It is also possible, with the aid of equation (2), to calculate the values of ϵ corresponding to the changes in thickness; the results of Fig. 6  were obtained in this way. From Fig. 6  and from Table I it is seen that each sample has practically the same voids-ratio under any given pressure, which indicates that the thickness of the sample has little effect upon the p/ϵ curve.

In his original research on the consolidation of clays, von Terzaghi found that the relation between voids-ratio and pressure could be expressed to a good approximation by the equation

where B is a constant and ϵ₁ denotes the voids-ratio under a pressure p₁ of unity. The experimental points lie on this equation, the values of the constants being ϵ₁ = 1·19 and B = 4·8.

The compressibility at any pressure p is readily found from equation (6), as

and the value of ϵ, which enables $α1+ϵ$ to be evaluated, can be taken from the p/ϵ curve or from equation (6).

In order to check the theoretical result that the value of μ at any time t varies inversely as d², the results of Fig. 5  have been replotted in Fig. 7 , showing the relation between μ and t/d². It will be seen from this that within close limits the theoretical result is confirmed, as was also the case with the other two pressure-increments.

In studying the general shape of the time-consolidation curves, one of the most convenient methods is to plot μ against $t,$ as first suggested by Drs. Gilboy and Taylor¹. The first half of the relation should then be represented by a straight line; and a typical $μ/t$ curve, given in Fig. 8 , shows that for the clay used in these experiments this is true.

From the slope of the linear portion of the $μ/t$ curve the value of c can be calculated, and the full theoretical curve can then be plotted from equation (4). Thus, in Fig. 8 , the slope of the linear portion is

and the value of d, which is one-half the mean thickness of the sample, is

From equation (5) it is known that

and hence, for equal slopes of the linear portions of the experimental and theoretical curves, the value of c is

In American laboratories it is usual to express the coefficient of consolidation in $(centimetres)2seconds.$ In these units the above value equals $8·0 × 10−5(centimetres)2seconds.$

The full theoretical curve can now be plotted, as shown dotted in Fig. 8 , and this is seen to deviate slightly from the experimental curve in the later stages of consolidation, as mentioned in the introduction. This deviation is known as the “ secondary time effect ” or as the “ secondary compression ” and, according to Drs. D. W. Taylor and W. Marchant, it “ may be of the nature of a plastic adjustment of the clay structure wherein all or practically all the shearing stresses are eventually dissipated¹.” The mechanism is not completely understood, and will not be discussed further, except to mention that in soils exhibiting considerable secondary compression a correction has to be applied in the calculation of c. The two principal methods of applying this correction are those in use at the Massachusetts Institute of Technology¹ and at Harvard University².

With London clay the secondary compression is small, and the simple method of calculating c, described above, can be used without serious error. The values of c determined by this method are given in Table II. They are about 10 per cent, lower than the effective values obtained by applying the corrections, but this difference is of little practical importance, particularly in view of the variations in the experimental values of c.

Within these variations, the Table shows that the coefficient of consolidation is independent of both pressure and thickness of sample. However, other oedometer tests on remoulded London clay, which have been carried out with pressures ranging up to 9 tons per square foot, indicate a small decrease in the coefficient at pressures greater than about 3 tons per square foot.

The results of numerous tests carried out on samples of London clay in its natural state reveal an agreement with theory of the same order as that described above. The value of the coefficient of consolidation is also of the same order, but the compressibility is, of course, very much lower.

The following conclusions are valid for London clay, and for the thicknesses of sample used :—

• the time/consolidation relation agrees with theory within reasonable practical limits,

• the rate of consolidation varies inversely as the square of the thickness of the sample,

• the coefficient of consolidation is a constant for the material, and within practical limits is independent of the pressure and thickness of the sample,

• the pressure/voids-ratio relation is independent of the thickness of the sample.