ARTICLES 577 



purposes it is sufficient. It would be possible, in cases where 

 more exact divisions were required, largely to increase the 

 number of the fractions, but the experimental work would 

 become very laborious, whether sedimentation or elutriation 

 was used. Oden, in LFpsala, has recently described an in- 

 genious and compact apparatus which will effect this detailed 

 subdivision. A soil suspension falls through water on to a 

 flat counterpoised plate and the time intervals for successive 

 small equal increments of weight are automatically recorded. 

 These data, when analysed mathematically, enable a distri- 

 bution curve to be constructed, with radii as abscissae, and the 

 proportion of particles lying between any two radii is given 

 in the usual way, by the area under the curve between these 

 specified radii. Oden's method will be of the utmost value 

 to soil physicists : it may be mentioned here that he has 

 already demonstrated that two samples of Globerina Ooze, 

 one from the Atlantic and the other from the Pacific, contained 

 no particles whatever between certain limits of radii — between 

 12 and 20/U- in the former, and i6 and 30yu, in the latter. This 

 obviously important deduction, which follows easily from 

 Odin's treatment, could not be obtained with any certainty 

 by other methods. 



Of the various fractions separated by ordinary mechanical 

 analysis, clay is the most important from the physical point of 

 view. It shows Brownian movement in water, it is flocculated 

 by acids and salts, deflocculated by dilute ammonia, and it 

 possesses considerable plasticity. In fact, the physical properties 

 of the vast majority of soils depend largely on the percentage 

 of the clay fraction present, which imposes its colloidal charac- 

 teristics on the soil mass. The earliest attempt to estirnate 

 what proportion of the clay fraction itself was in the colloidal 

 condition was made by Schloesing in 1874. By evaporating 

 the water which had stood for many weeks in contact with clay, 

 he obtained a very plastic material of a horny nature, which 

 was translucent in layers up to i mm, in thickness. This sub- 

 stance constituted only about 2 per cent, by weight of the 

 original clay. Recently Tempany, by measuring the shrinkage 

 of blocks of moist compacted soil, has shown that the percentage 

 of colloidal clay is in excess of 2 per cent., and that its amount 

 varies from one clay to another. It is probable that the 

 methods of Tempany and Schloesing give respectively the upper 

 and lower limits, and that the actual percentage lies between 

 them. Certain of the soils used by Tempany were very difficult 

 to keep in good cultivation, as they became sticky in wet 

 weather, and dried to hard clods in the summer. These soils 

 did not necessarily contain the greatest amount of clay, but, 

 of the clay present, a greater percentage was in the colloidal 



