ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 37 
temperature coefficient of viscosity determines very largely the 
value of the temperature coefficient of electro-osmose, is strengthened 
by experiments of Coehn and Raydt,*t Cameron and Oettinger,”® and 
of Briggs, Bennett and Pierson.** 
On the other hand Cruse” found an apparent maximum in the 
rate of electro-osmose, this maximum occurring between 35° and 
40° C. Briggs, Bennett, and Pierson have shown that Cruse’s 
maximum was probably due to the diaphragm not being in equi- 
librium with the liquid phase. They have carried out determination 
of the temperature coefficient and find that with diaphragms of 
asbestos, cellulose (filter paper) and carborundum, the rate of flow 
increases with rising temperature slightly less rapidly than the 
viscosity decreases; the rate for a given diaphragm at constant 
temperature is also very exactly proportional to the applied external 
potential. It is only fair to point out, however, that since adsorption 
varies with the temperature, and since the temperature coefficients for 
anion and cation are not necessarily equal, « may also vary with the 
temperature and might even change its sign. In the cases just 
discussed, « is probably very nearly constant. 
6. Related Phenomena.—Cataphoresis is the reverse of electrical 
endosmose. One is dealing here with a mobile or “floating” dia- 
phragm. The fundamental theory is without question the same ; 
such differences as appear to exist between ihe two processes being 
due, in all likelihood, to flocculation complicating the phenomena of 
cataphoresis. The particles are positive if they migrate to the cathode, 
negative if they migrate to the anode. Acid-alkali reversals are 
known, and strongly adsorbed ions determine the charge on the 
particles. As the subject constitutes a separate report in this series, 
it will not be considered further at this time. 
Perrin® has pointed out that the curious Bose-Guillaume phe- 
nomena is a special case of Quincke’s “ diaphragm currents,” and is an 
electro-osmotic phenomenon. Guillaume’s data®® are evidence in 
favour of the adsorption theory. Electro-stenolysis is another phe- 
nomenon dependent upon contact electrification. 
When certain aqueous solutions are separated from water by a 
porous plate of porcelain, for instance, a slight but measurable osmosis 
is often observed. In some instances the osmotic flow occurs from 
the solution into the water and not from the water into the solution 
as one would expect it todo. Such apparently abnormal osmosis is 
termed “negative osmose.”® 
According to Bartell the most plausible way of accounting for 
osmosis of this type is to assume that it is caused by the “ polar- 
ization”’ of the membrane. The membrane being in contact with 
different liquids, the adsorption potential will be different for each 
4 Drude’s Ann., 30, 797 (1909). 
55 Phil. Mag. (6) 18, 586 (1909). 
58 Jour. Phys. Chem. 22, 256 (1918). 
7 Phys. Zeit., &, 201 (1905). 
58 Comptes rendus, 147, 55 (1908). 
59 Thid. 14'7, 53 (1908). 
® Bartell, Jour. Am. Chem, Soc. 36, 646 (1914) ; Bartell and Hocker, Jour. am. 
Chem. Soc., 38, 1029 (1916) ; Freundlich, Zeit. Kolloidchemie, 18, 11 (1916) ; Girard, 
| Comptes rendus, 146, 927 (1908) ; 151, 99 (1910) et seq. 
. 20895 B3 
