28 BULLETIN 1311, U. S. DEPARTMENT OF AGRICULTURE 
after freezing and thawing, gave 86 parts per million of total solids, 
while the untreated soil gave 76 parts per million. 
By cataphoresis. — Ormandy, as reported by Briggs ('/), considers 
that an " osmose machine,'- which is essentially a cataphoresis cell 
with a slowly revolving anode, will purify ceramic clays by separat- 
ing part of the iron, which goes to the cathode, the purified clay 
being deposited on the anode. This claim that the clay can be puri- 
fied by a separation of the iron has been questioned by Briggs (7), 
who points out that it is by no means certain that the iron oxide and 
particles of the pure clay substance will bear opposite charges in the 
suspension, as assumed by Ormandy. Briggs further contends that 
" if the iron oxide is absorbed by the kaolin the electrical treatment 
can not remove an appreciable amount of iron from the clay." 
Ormandy reports that out of 25 clays worked over by the osmose 
machine, 15 contained less iron after purification and 8 contained 
more. 
This partial purification reported by Ormandy led to the belief 
that a separation of soil colloidal matter into simpler constituents 
might be made by cataphoresis. For this purpose very dilute sus- 
pensions of the colloidal matter from seven soils, Nos. 3, 6, 11, 12, 
20, 37, 41, and a very pure and finely ground kaolinite were subjected 
to cataphoresis for several days. All the colloidal material appeared 
to be made up of negatively charged particles. No particle of any 
suspension appeared to be positively charged. 
The colloidal materials tested represent a wide range of chemical 
composition and other properties, and it is doubtful if the other soil 
colloids could be fractionated b}^ cataphoresis, since all the colloidal 
particles in a soil suspension appear to have the same charge. 
Conclusions from attempted fractionations. — The attempts to sepa- 
rate the colloidal material into fractions of different compositions 
were unsuccessful and showed nothing concerning the constitution of 
the colloidal matter, except that if there are different compounds 
present, the mixture must be very intimate. 
In order to make a physical separation of the colloidal matter into 
fractions of different composition, it would seem necessary that sepa- 
rate particles of different composition exist in the mixture to be sepa- 
rated. This is hardly to be expected. Certain possible components 
of the colloidal mixture, such as aluminum oxide a'hd silicic acid, 
would doubtless bear opposite charges under the conditions of sus- 
pension and therefore coalesce to one particle in the mixture. The 
negative results of the attempted separations support the idea that 
separate particles of radically different composition do not exist in 
the suspension. 
ACTION OF SOLVENTS ON COLLOIDAL MATERIAL 
Water. — The various constituents of the colloidal matter presum- 
ably differ considerably in solubility in various solvents, as indi- 
cated by the work of Schloesing (38) and Van Bemmelen (3). Ic 
would seem that some progress might be made in evaluating different 
compounds that may be present by the use of certain solvents. In 
tli is connection it became of interest to see if there was a differential 
solubility of the constituents in water. This seems probable, for 
drainage waters contain relatively large amounts of calcium, sodium, 
potassium, magnesium, sulphur, and chlorine and relatively little 
