312 Mr. H. A. M c Taggart on the 



and N0 3 . In the other two cases hydrolysis produce a 

 supply or! H-ions, to which the higher activity of the Al-salt 

 may be ascribed. One would expect the Cr-salt for the same 

 reason to be more active than La, but its effect is even less. 

 Similar discrepancies occur in the stability of. these metals in 

 colloidal form (Biltz, Bed. Ber. xxxv. p. 4434), and no 

 satisfactory explanation seems possible at present. 



Of the two tetravalent nitrates Th and Zr the first is 

 unusually active, the merest trace of the salt being sufficient 

 to reverse the charge. Both salts hydrolyse and their 

 colloidal hydroxides are prepared by dialysis. Thorium is 

 remarkable also for its stability as a colloid, being much more 

 stable in the presence of added electrolytes than Zr (Biltz, 

 loc. cit.). In endosmose experiments using water against 

 glass and quartz, Elissafoff {Zeit. Phys. Chem. lxxix. p. 385, 

 1912) finds that Th(N0 3 ) 4 has the same high charging power. 

 This activity cannot be due to the presence of H-ions resulting 

 from hydrolysis altogether, because the conductivity of the 

 solution is relatively small. It has a large atomic weight, a 

 fact worth noting in connexion with similar observations on 

 the activity of heavy metals in endosmose experiments. 



The chlorides of platinum and tin and the oxide of chromium 

 evidently produce ions whose charges are small. Simple 

 ions are not formed in these cases — a necessary condition 

 apparently for the production of the greatest effect on the 

 surface electrification. 



The two negative ions, FeCy 6 *'° and FeCy 6 — , follow the 

 valency rule that the ion with a larger charge is more active 

 than one with a smaller charge. Similar results for positively 

 charged copper in colloidal solution have been obtained by 

 Burton (Phil. Mag. vol. xvii. no. 100, April 1909) who found 

 the charge to be reversed by K 3 P0 4 and K 3 (FeCy 6 ). 



In a general way the valency determines the charging 

 power of an ion, but it is not the only factor apparently 

 which has an influence. The greater acti\ity of the ions of 

 the heavy metals has been mentioned. The complexity of 

 the ion also is associated with peculiarities in its activity. 

 Mines (loc. cit.) has examined the action on colloidal substances 

 of the complex trivalent ion [Co(] s sH 3 ) 6 ] in Luteo-cobalt 

 chloride and of simple trivalent ions. Some colloids are 

 sensitive to both complex and simple ions, and some only to 

 simple ones. An emulsion of olive-oil in water differentiates 

 between the two, the simple ion being 170 times more 

 powerful in precipitating the oil. 



From these experiments with a rotating cell, it would seem 



