118 The Petrography and Genesis of Sediments 



The clay washings in the large separator funnel were allowed to settle 

 for about half an hour, the settlings tapped and rewashed for any sand or 

 silt that might have escaped the first washing. The amount, as indicated 

 above, was usually very small. This method of separating sand and 

 " clay " is in principal entirely similar to the method of the Bureau of 

 Soils of the U. S. Department of Agriculture. 1 The use of the centrifuge 

 by the Bureau of Soils merely hastens settling. Their method differs 

 mainly in having definite size limits for the finer portions of " silt " and 

 " clay." But the analysts of the Bureau of Soils themselves recognize that 

 a perfect separation is never attained and that it is indeed theoretically 

 possible only if all the particles treated have the same density and shape. 

 But if other conditions remain constant the same result is attained by 

 allowing the particles to settle a definite length of time through a fixed 

 distance, so that theoretically the method of Thoulet, as used, gives the 

 same results, even though time of settling instead of prevailing size of 

 particles settled is used as the determining factor. 



In general, it must be said, and is admitted by all students of sediments, 

 that all such mechanical methods of separating sand and " clay," while 

 they allow valuable comparison, are, from a scientific standpoint, still 

 most unsatisfactory. It is now generally believed that the colloidal state, 

 in which true clay may be assumed to be, is merely a certain state of sub- 

 division between fairly definite limits (lOO^/u. to 10/x//.) in a continuous 

 series from grains visible to the unaided eye to molecular solution. If this 

 is so, then any separation of what might be called true clay, even if it were 

 mechanically possible, would still be somewhat arbitrary. Moreover, 

 there is some reason to doubt that in a natural sediment there actually 

 exists such a continuous series rather than a mixture of certain definite 

 consituents or groups of constituents each with its own size limits, the 

 limits overlapping more or less. 



The ideal solution of the problem would be to establish a curve showing 

 the rate at which the settling of the constituents of a given sediment 

 progresses. That different constituents can be differentiated in the finest 



1 V. S. Dept. Agric. Bureau of Soils, Bull. No. 24, 1904; Bull. No. 84, 1912. 



