In spite of our present very advanced analytical capabilities, our present ability to predict actual field behavior is, in many cases, woefully inadequate. Four cases are described to illustrate one of the reasons why actual and predicted performances are sometimes so widely divergent—namely, failure to understand how a soil might respond over time to changed conditions. The four cases are: (1) The aging of quick clay after sampling, in which the remolded strength increases in samples maintained at constant water content; (2) time effects in freshly densified or deposited sand, in which natural sand deposits can lose strength if disturbed but regain strength over time periods of weeks to months; (3) apparently sound lime‐stabilized soil that swells and disintegrates starting a few years after construction; and (4) the failure of excess pore pressures to dissipate as predicted during the consolidation of soft clays. Study of each of these problems has led to an understanding of the responsible phenomena and to technical conclusions useful in geotechnical practice. Collectively, they illustrate that soils are not inert materials, but that they can, and often do, change with time and changes in environmental conditions. Soil mechanics, in either conventional or advanced forms, cannot account for three of the four cases described. All conceivable physicochemical and biological changes must be considered in any case to explain the unusual behavior. It is essential to expect the unexpected and to deal with soils as they are, not as we might wish them to be.

A new, simple and approximate method of analysis for tapered piles under vertical vibration is presented. The method of analysis accounts for soilpile interaction in a relatively simple manner. Since cross‐section and dynamic soil reaction per unit length of tapered piles vary with depth, the solution for dynamic response is obtained by a finite difference method. The analytical solutions agree satisfactorily with reported experimental results. The dynamic response and the results of a parametric study of tapered concrete piles are presented in non‐dimensional form. With an increase of angle of taper, resonant amplitude decreases and resonant frequency increases. In general, dimensionless stiffness and damping parameters are frequency dependent and also vary with the slenderness ratio and Poisson's ratio of the soil medium. In all cases, however, stiffness increases and damping decreases with increase of angle of taper of piles.

In this work, an analytic expression has been developed that enables the shear strength to be determined using the empirical failure criterion developed by Hoek and Brown. Applying this new hypothesis, along with elasticity theory and basic mathematical concepts for obtaining the failure envelope of a family of failure circles, the shear strength has been determined to be such that ${\tau }_{\alpha }=f\left({\sigma }_{\alpha }\text{,}{\sigma }_c\text{,}m\text{,}s\right)\text{,}$ where ${\sigma }_{\alpha }=\mathrm{the}$ normal stress on the discontinuity, ${\sigma }_c$ the uniaxial compressive strength of the intact rock and the parameters m and s are constants dependent upon the properties of the rock. An expression of this type will help us develop new methods of analysis of slope stability and support of galleries, design of anchor support systems in rock masses, estimation of the skin friction of poor quality rock in driven pile foundations, as well as in other innumerable related areas of geotechnology.

In this paper the writers first present experimental results obtained for the distribution of the active stresses due to a sand backfill behind a rigid wall rotating about the top. The experimental evidence shows that the stress distribution is nonlinear and that, due to arching, the stress near the top of the wall increases beyond the level of the at‐rest stress. Consequently, the point of application of the lateral thrust is much higher than one‐third from the wall base. Arching effects increase with increasing soil density. Secondly, comparisons of the active earth pressure distributions are made for three different wall movement modes: (1) Rotation about top; (2) rotation about heel; and (3) translation. Total active resultant forces and the points of application of these forces are summarized.

To support a theory proposed to explain the mechanism causing blast‐induced liquefaction, a series of high pressure undrained isotropic compression tests have been performed on Monterey No. 0 and Eniwetok sands. These tests show that sand can be liquefied (in the sense of 100% pore pressure rise) by a single cycle of isotropic compressive stress applied under quasistatic, undrained conditions. Eniwetok sand was found to be much easier to liquefy than Monterey sand. Initial dry density was not a significant factor in the generation of residual excess pore pressure for either sand. Drained isotropic compression tests and membrane penetration tests were also performed on the sands. These tests were used to determine the magnitude of system compliance and to allow calculations of predicted undrained behavior based on drained material properties. System compliance was shown to inhibit the development of liquefaction. Predicted undrained behavior compared favorably with test results, although additional work is needed. A series of grain crushing experiments show that Eniwetok sand crushes more easily than Monterey sand. This may account for the difference in pore pressure response.

A computer‐based analysis of the runout dynamics of the Madison Canyon rockslide of August 17, 1959, is reported. Three levels of computer representation of the flow are considered. First is a representation of the actual canyon profile, where the model assumes incompressible linear viscous flow. Second is a horizontal profile in which gravitational forces are input by components, and again the flow is approximated by a single viscosity model. Third is a horizontal profile, for which a two viscosity model approaching a Bingham material representation is used. Results of each model are compared to actual slide runout, debris distribution, and maximum speed of the flow. The biviscous model, which depicts an active basal flow layer, also has physical characteristics consistent with measurements made of the Madison Canyon slide following the event.

The dynamic impedances of a vertically or torsionally excited viscoelastic soil layer with a thin disturbed zone around the edge of a central circular hole are evaluated and compared, over a range of parameters, with those obtained by a previously proposed, simpler analysis in which the mass of the disturbed zone is neglected. It is shown that the results of the two approaches may differ significantly, and that the simpler solution should be used only with the greatest possible care.

In recent years, attempts have been made to correlate standard penetration blow counts with cone penetration tip resistance readings in order to utilize the large data base and experience of standard penetration tests to interpret the results of cone penetration soundings. As part of a liquefaction-susceptibility study on a site consisting of a laterally-extensive hydraulic sand fill over natural sands, correlation data were obtained from side-by-side standard penetration and cone penetration tests. The standard penetration tests were corrected on a standard uniform energy basis and the cone penetration tests were made with an electronic cone. It was concluded that the correlation between the mean grain size of the sand and the ratio of the cone tip resistance to standard penetration test blow counts vary significantly with the presence of fines (material smaller than a US Standard #=200 size sieve). It is suggested that this variation be considered when making evaluations of settlement, stability, and liquefaction analyses for soil profiles explored primarily with a cone penetrometer.

Several formulae and charts have been published for correcting the results of the standard penetration test (SPT) performed in sands for the effects of overburden pressure. However, depending on which published correction factor is used, very different interpretations may result. This paper attempts to clarify and resolve some of these differences, and proposes a simple correction factor as a tentative standard, contingent on data from future research.