Hydroplastic gravity core no evidence for shortening occurred in the top approximately 

 50 cm. It may be possible to take longer gravity cores with a 100 percent gross re- 

 covery ratio using even better engineering design. 



Sampling disturbance — Results of tests on varied clays led Hvorslev (1949, 

 p. 105-109) to recommend that long samplers in cohesive sediments should have an 

 inside clearance ratio of 0.75 to 1 .5 percent, an outside clearance ratio of less than 

 3 percent, and an area ratio less than 10 percent; a greater area ratio was considered 

 permissible if the sampler had a stationary piston and/or a very small cutting edge 

 angle . 



It was not possible to easily test Hvorslev's recommendations on dimensions of 

 corers until the past decade, when In-place strength tests made by the vane borer 

 (bibliography in Osterberg, 1957) were compared to different types of samplers. Two 

 important papers by Jakobson (1954) and Kallsteius (1958) resulted from extensive tests 

 conducted on relatively homogeneous post-glacial clays by the Royal Swedish Geo- 

 technical Institue. These tests in general confirmed Hvorslev's recommendations and 

 showed that piston samplers obtained undisturbed samples only if properly engineered. 

 Kallstenius (1958, p. 67) emphasized the importance of a small cutting-edge angle. 

 Most corers used by oceanographers have large edge angles, in Norway, Bjerrum 

 (1954a, p. 55) reported that a redesigned thin-wall piston sampler having a small area 

 ratio successfully obtained undisturbed samples in sensitive clays. 



Photographs or drawings of strata in piston cores (Arrhenius, 1952, appendix; 

 Ericson and Wollin, 1956, Fig. 4; Sykes, 1960, Figs. 28-30; and others) amply show 

 that submarine piston cores are not undisturbed; inspection of clearance and area ratios 

 (Table 3) clearly infers why. 



Small diameter, poorly engineered (for undisturbed sampling) gravity cores prob- 

 ably show disturbance resulting from core shortening of Type II in Figure 5. It is 

 noteworthy that better engineered, larger-diameter gravity corers do not show core 

 shortening near the surface (Richards and Keller, in press), and hence take more nearly 

 undisturbed samples. The Hvorslev -Stetson (1946) gravity corer also may take little- 

 disturbed samples because of its good design (Table 3). 



A detailed analysis of disturbance caused by sampler shape has been written by 

 Kallstenius (1958, p. 40-66), should the reader desire additional information on how 

 disturbance is caused. 



Summary — It has not been demonstrated that piston cores taken by oceanogra- 

 phers navegross recovery ratios of 100 percent. Tests and experience reported by 

 Scandinavian soil engineers suggests that sample disturbance, caused by excessive 

 clearance and area ratios and large taper angle of the cutting edge of these corers, 

 could be reduced by applying modern engineering techniques. 



15 



