gravel, and rock are difficult to core, and often poor samples or none at 

 all are obtained. For hard coarse materials (gravel, sand) a scoopfish 

 or bucket dredge nnay obtain the most representative sanaple. 



B, Generally less representative than cores are dredge, trawl, 

 grab or snapper, and scoopfish samples, even -when carefully analyzed 

 by conscientious and competent persons. The grab and snapper samples 

 usually destroy the structure of the sediment and often bias the grain 

 size distribution toward coarser sizes by winnowing or washing in the 

 sampling process. Trawls and dredges also nnay bias the size distri- 

 bution by preferential selection of certain sizes. However, they provide 

 composite samples of the submarine bottom nnaterial over an area 

 or along a track. While such a composite sample is not representative 

 of the material at any specific spot, it probably is more representative 

 of the range of material in a region than one or even several cores. 

 Therefore, one composite sample might better describe the bottonn 

 materials of a given area than one core sample, but ten core samples 

 would describe more accurately the materials of the same area than 

 would ten composite samples. 



C. The poorest data and the most numerous are abbreviated nota- 

 tions of types and consistency of bottom materials which appear on 

 boat sheets, compilation sheets, and printed nautical charts. Most 

 of them are cursory identifications of small samples, often only a few 

 grains,- brought up on sounding leads in the course of hydrographic 

 surveys. The identifications of these inadequate samples are made 

 routinely by enlisted naval personnel who usually have little or no 



scientific training in sedimentary petrology or mineralogy. Navigational 

 positioning errors are compounded by cartographic compilation and 

 printing errors. Although some of the notations are highly reliable 

 and accurate abbreviations of field descriptions of samples from the 

 other two groups, they usually cannot be distinguished from the 

 majority of less reliable data and must be classed with them. The total 

 effect of all of these possible sources of errors significantly reduces 

 the reliability of bottom sediment charts based solely upon nautical 

 chart notations. The charts become much more reliable when chart 

 notations are substantiated by numerous and well-distributed core and 

 grab samples. 



The bottom sediment data collection maintained by the U. S. Naval 

 Oceanographic Office (apart from chart compilations) consists almost 

 entirely of data of the first two classes. Data in the third class are 

 included only when reported in professional journals (such as Annalen 

 der Hydrographie, Deutsches Hydrographisches Zeitschrift, and Annales 



