30 REPORTS ON THE STATE OF SCIENCE.—1918. 
solution, the point was really established with remarkable clearness 
by Parker. a 
Perrin’s contribution!® to what was still an obscure phenomenon, 
did, however, constitute a great step forward. He devised an 
apparatus which enabled him to employ many different powdered 
substances in the form of pervious diaphragms. Turning his atten- 
tion to the effect of dissolved substances and using extremely dilute 
solutions (never more concentrated than N/50 and usually about 
N/500), he concluded first that electrolytes alone among solutes 
influenced the course of electrical endosmose and that the ions had 
a specific effect. Hydrogen and hydroxy] ions proved particularly 
active. With a diaphragm of insoluble chromic chloride, for example, 
dilute alkalies flowed to the cathode whilst dilute acids flowed to the 
anode—a clear cut reversal. Similar acid alkali reversals were found 
with alumina, carborundum, sulphur, gelatine, graphite, naphthalene, 
etc., although none was observed with cotton wool, glass and 
iodoform. Whenever reversal did occur, there was a certain 
hydrogen ion concentration at which no flow at all occurred, this 
point corresponding approximately to an isoelectric condition of the 
diaphragm. : 
Since flow to the cathode indicated an electronegative diaphragm, 
whilst flow to the anode pointed to an electropositive one, Perrin 
concluded as follows :— 
‘The electric potential of any surface whatever in aqueous 
solution is invariably increased [made more positive or less negative | 
by the addition of a monobasic acid and is invariably lowered 
[made more negative or less positive] by the addition of a non- 
acid base.’ 
Perrin considered next a large number of other ions. He found 
that, excepting hydrogen and hydroxyl, none of the common 
univalent ions was of much influence on electrical endosmose. 
Polyvalent ions were more active. Briefly, the results may be sum- 
marized in the following statement, sometimes referred to as 
Perrin’s valence rule, which one should compare with the Schulze- 
Hardy rule of flocculation : 
Every diaphragm tends to become charged nee y in an acid 
solution and negatively in an alkaline one. Every ion of unlike 
sign tends to neutralize the charge on the diaphragm and this 
tendency increases rapidly with the valence. 
Perrin’s rule has been extended to alcoholic solutions: by 
Baudouin”, incompletely to liquid ammonia by Ascoli?! and by 
Guillaume” to the so-called Bose-Guillaume phenomenon. 
Morse and Horn?’ have made use of electrical endosmose to 
remove air from the pores of porous cups, before impregnating the 
Johns Hopkins Dissertations, 31, 23 (1901); Cf. Briggs, Jour. Phys. Chem. 
Q1, 235 (1917). 
18 Jour. Chim. Phys. 2, 601 (1904). 
19 Cf., however, Bethe and Toropoff, Zeit. Phys. Chem. 89, 597 (1915). 
A Comptes rendus, 138, 898 (1904). 
21 Comptes rendus, 137, 1253 (1903). 
» 22 Thid. 147, 53 (1908) ; Cf. Briggs: Jour. Phys. Chem. 9], 215 (1917). 
23 Amer. Chem. Jour., 26, 801 (1901). 
