ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 31 
latter with copper ferrocyanide. This led Frazer and Holmes* to 
carry out experiments with an elaborate electro-osmometer, in which 
they noticed the usual acid-alkali reversals and concluded that the 
effect of the ions bore some relation to their migration velocities. 
Coehn® has also carried out some interesting electro osmotic 
measurements with a Pukall filter and solutions of H.SO,4, CuS0Ox, 
HNOs, acetic acid, alkali metal sulphates, alkali and heavy metal 
nitrates. 
Barratt and Harris” in some excellent work with diaphragms of 
gelatine, agar and parchment, have confirmed in the main the valence 
rule of Perrin. Their data suggest the problem of finding out what 
effect the salts of Hofmeister’s series would have on endosmose 
through gelatine and other albumenoids. Recently this question has 
been investigated by Bethe and Toropoff?’ and although the data are 
incomplete, there is reason to believe that the lyotrope series really 
does play an important part. 
An elaborate study of electric endosmose in neutral solutions has 
been completed by Elissafoff?* in Freundlich’s laboratory. The pro- 
cedure was based on Lemstrém’s?? method of measuring endosmetic 
movements in a single capillary without the direct application of 
electrodes. Elissafoff’s data show that Perrin’s valence rule holds only 
toalimited extent. So far as the alkali metal and alkaline earth metal 
cations are concerned the valency rule holds satisfactorily,*° but fails 
in the case of heavy metal and organic cations. For example, certain 
univalent organic cations appeared quite as effective in neutralizing 
the charge (see the second part of Perrin’s rule) on negative glass or 
quartz as did the divalent light metal cations. Silver as ion belonged 
to the divalent metals; mercuric chloride was as effective as 
aluminum sulphate and far more so than chlorides of the divalent 
alkaline earths. That the anion is also active was demonstrated by 
experiments with potassium benzoate and sodium picrate, these 
anions restraining the tendency of the potassium or sodium to 
neutralize the negative charge on the walls of the capillary. 
No reversal by acids was observed, but the strength of the 
solutions employed was never greater than one ten-thousandth 
normal. With certain salts, however, reversals were obtained. 
Against solutions of thorium nitrate of sufficiently high concentra- 
tion the glass or quartz became electropositive and a similar reversal 
occurred with solutions of methyl violet, a basic dye. One should 
compare this with what Larguier des Bancels*! found with wool and 
basis dyes, where the latter showed a striking tendency to make the 
wool positive. 
24 Amer. Chem. Jour., 40, 319 (1908); H. N. Holmes, Jissertation (Johns 
Hopkins, 1907). 
25 Zeit. Elektrochemic, 1G, 586 (1910). 
26 Zeit, Blektrochemie, 18, 221 (1912) ; Biochemical Jour., 6, 315 (1912). 
27 Loe. cit. 
% Zeit. Phys. Chem., 9, 385 (1912). 
*9 Drude’s Ann., 5, 729 (1901). 
30 Freundlich and Elissafoff have employed electro-osmosis to estimate the 
valency of radium. The results agree with the assumption that radium is a 
divalent metal belonging to the alkaline earth group. Phys. Zeit., 14, 1052 (1913). 
31 Comptes rendus, 149, 316 (1909). 
