4 BULLETIN 1193, U. S. DEPARTMENT OF AGRICULTURE. 
showed that his argile colloidale was probably very similar to the 
colloid we separated by the supercentrifuge; at least his largest 
colloidal clay particles could not well have been smaller than the 
E articles passing the supercentrifuge. Schloesing evidently based 
is conclusions concerning the small quantities of colloidal clay 
present in soils on the quantities he succeeded in isolating. His 
error, which has since been made by several other investigators, lay 
in assuming that his means of extraction and purification were sub- 
stantially quantitative. As a matter of fact, he probably dispersed 
only part of the colloidal matter in his preliminary treatment of the 
soil and then lost colloid in the residues from which he repeatedly 
decanted. 
About the same time as Schloesing. Hilgard (16) separated a 
" colloidal clay 7 ' from the soil. This colloidal clay he believed con- 
tained a "true clay substance," with a chemical composition similar 
to kaolinite, which was responsible for the plastic and adhesive 
properties of the soil. Hilgard classified as colloidal clay those 
particles that would "fail to settle in the course of 24 hours through 
a column of pure water 8 inches high'"' {15). He stated (15 p. 333) 
that the percentage of this material "seems rarely to reach 75 in the 
purest natural clays. 40 to 47 in the heaviest of clay soils, and 10 to 
20 in ordinary loams.'" 
Apparently Hilgard believed that this fraction would contain, 
besides "the true clay substance," other colloidal material, such as 
ferric, silicic, and aluminic hydrates, which he assumed would 
impart no plasticity to soils. The material he obtained, however, 
seems to have had the same properties as that of Schloesing. Hil- 
gard's method of extraction was possibly more thorough than that of 
Schloesing, but he doubtless included in his colloidal clay, particles 
that Schloesing graded out of his argile colloidale. 
The very painstaking method of mechanical analysis reported by 
W. R. Williams in 1895 (39) was evidently somewhat more effective 
than that of Hilgard in extracting the colloidal matter of soils, in 
that the fractions of coarser soil particles were more thoroughly 
worked to rid them of colloidal material, and the method probably 
gave a more complete separation of colloidal and noncolloidal 
material. 
The " schlamm " fraction secured by Williams, similar to the " clay " 
fraction of Hilgard. contained only those particles which would 
remain in suspension for 24 hours in a column of water 10 centimeters 
high. Williams stated that these particles, which were less than 1.5 
microns in diameter ; had very different properties from the larger 
particles. The schlamm material gave a good suspension of formless 
particles that showed an active Brownian movement. On drying, 
the material shrank greatly, becoming a horny mass of conchoidal 
fracture that would adhere strongly to the tongue. Williams 
attributed the coherence, plasticity, and adsorptive power of soils to 
the schlamm material, and stated that this material constituted 1.5 
per cent to 40 per cent of the soil (39 p. 301). 
The materials isolated by Schloesing, Hilgard, and Williams were 
evidently all about the same and were chiefly colloidal. Apparently, 
Ehrenberg (8 p. 99) does not recognize this, as in his recent com- 
pendium of colloidal phenomena in soils he agrees with Schloesing that 
soils contain only 0.5 per cent to 1.5 per cent of colloidal clay. Ehren- 
