caused a 10 percent reduction In shear strength compared to In-place tests. Improved 

 piston samplers carefully designed in light of recommendation by Hvorslev (1949) 

 have proven themselves In Norway (BJerrum, 1954a, p. 55; Void, 1956) and In 

 Sweden (Kallstentus, 1958, p. 72-73), as previously mentioned . Initial success v/Ith 

 the Hydroplastlc corer (Richards and Keller, in press) suggests that better engineered 

 gravity and piston corers may provide a partial solution In submarine investigations. 

 A still better solution Is in-place strength testing. This already has been accomplished 

 In the shallow-water, marine environment (Moore, personal communication), but in- 

 place testing in deep water has yet to be attempted. 



One means of estimating the likelihood of disturbance is by measurement of sensi- 

 tivity; the greater the sensitivity the larger the loss of strength In the completely dis- 

 turbed or remolded condition. Skempton and Northey (1952, p. 31) have classified 

 clay-size (silt-size particles also show sensitivity, p. 33) sediments according to sensi- 

 tivity. Their classification has been extended by Rosenqvist (1953, p. 195), whose 

 system is shown In Table 5, together with the percentage loss of strength upon remold- 

 ing an "undisturbed" sample computed by me. 



Sensitivities of core samples are given in Table 6 and Figure 22. Most values 

 are In the range of 2 to 6, medium to very sensitive, while three from Area H have 

 sensitivities between 17 and 26, indicating that the samples tested are medium quick. 

 (The word "quick" in soil mechanics denotes a fluid consistency In the remolded state 

 without much strength). Sensitivities resulting from vane tests In Core H 12 (Fig. 20) 

 are higher than those from compression tests as a result of lower values of remolded 

 cohesion measured by the vane tests. Although the in-place sensitivities of deep-sea 

 sediments ore unknown, the presence of medium to very sensitive sediments, of medium 

 to high water content (Richards, in preparation). In most of the cores is an indication 

 that In-place sediments may be even more sensitive; more loss of strength may have 

 occurred In sampling than suspected. 



Although values of shear strength presented are conservative by an unknown 

 amount compared to in-place strength, because a quantitative estimate of the reduc- 

 tion of in-place strength cannot be made with the Information available, values 

 reported are considered sufficiently reliable for engineering use at the present time. 



42 



