120 The Petrography and Genesis of Sediments 



that hot water dissolved a small portion of it, partly salts, partly a tough, 

 almost white, colloidal substance, so that in the later analyses the whole 

 quantity was evaporated down on a steam bath. The dried clay was 

 scraped out of the dish with a steel spatula, a process which always 

 involved some loss, partly from a small residue that adhered, partly from 

 dust that was carried away. The clay was then dried for eight hours or 

 more at about 105° C, cooled in a desiccator and weighed as rapidly as 

 possible. 



The sand separated from clay and silt was air-dried, weighed and then 

 passed through a series of sives made of bolting cloth with approximately 

 30 (28), 60, 100 (97), and 200 meshes, respectively, to the inch. 1 



Following, according to Thoulet's observations, 2 are the minimum sizes 



of the materials held back by the different sieves : 



= Coarse sand. 



30 0.89 mm. 



= Medium sand. 



60 0.45 " 



= Fine sand. 



100 0.26 " 



= Very fine sand. 



200 0.04 " 



= Extra fine sand. 



Even this simple process of sieving is not quantitatively absolute which, 



as indicated above, is one of the reasons for preferring the elutriation 



method. The two causes are : most important of all that the grains are 



not round ; a minor factor that the meshes, especially in the finer bolting 



cloths, are not uniform. As a result of the irregular form of the grains, 



very long grains with a short diameter less than the mesh opening will 



pass, and with prolonged shaking very many of them. The duration of the 



sieving is, therefore, a matter of accommodation based largely on personal 



judgment and experience. The procedure was to stop when the grains 



that came through were predominantly elongated. But this stage will be 



1 The figures in parentheses are the given meshes, according to trade num- 

 bering, which were the nearest that could he obtained. The actual mesh, ac- 

 cording to measurement, is still somewhat different, in most cases fewer 

 meshes per inch or larger openings. Professor Thoulet was, however, good 

 enough to assure the author that these were quite accurate enough. 



2 Thoulet, J., Precis d'analyse (op. cit.), p. 64. 



