84 



Dr. W. M. Bayliss. 



[Apr. 7, 



must be present, otherwise the red salt of the dye with this alkali is formed 

 at once. The adsorption compound, if formed at all, only exists for an 

 infinitesimally short time, owing to the rapidity of the chemical reaction. 



In order to obtain as large a relative surface of the hydroxide as possible 

 I have made various hydroxides in colloidal solution, prepared by dialysis 

 of solutions of salts which are hydrolytically dissociated. Ferric chloride, 

 aluminium acetate, zirconium and thorium nitrates have been treated in 

 this way. With ferric hydroxide, although the result of the experiment is 

 quite distinct, the change of colour on heating is not so obvious as with a 

 colourless hydroxide. Aluminium hydroxide is good, but is unstable when 

 sufficiently dialysed. The best of all those with which I have worked are 

 the colloidal hydroxides of zirconium and of thorium, which are beautifully 

 clear and colourless solutions. The clearness is, of course, an indication of 

 the minute size of the suspended particles. Like all solutions prepared in 

 the way described, they still contain, even after prolonged dialysis, traces 

 of the original acid. If this is present in too large a proportion, no red salt 

 is formed even on heating the adsorption compound. This fact was shown 

 in a striking way in my first preparation of zirconium hydroxide, which had 

 been insufficiently dialysed. In this case, although the adsorption com- 

 pound was duly precipitated, it did not become red on heating. When the 

 adsorption compound was suspended in water and subjected to further 

 dialysis, it was noticed to be turning slightly reddish at room temperature ; 

 on boiling, the change to the red salt was immediate. The compound with 

 thorium hydroxide seems to require heating for a longer time before com- 

 bination occurs than do the others ; but this may be merely owing to the 

 presence of more acid in the particular preparation. 



As is usual in the case of mutual precipitation of colloids, the readiness 

 with which the precipitate falls depends on the rate at which the one 

 colloid is added to the other.* If the rate is too slow, or the solutions are 

 too dilute, the adsorption compound does not deposit but remains in suspen- 

 sion as a complex colloid. Moreover, a trace of a protective colloid, such as 

 gelatin, is sufficient to prevent precipitation. 



An analogous case to the one just described was met with by Van 

 Bemmelen.-f- If barium hydroxide solution be added to colloidal silica, a 

 white precipitate falls, which is found to contain both barium hydroxide 

 and silicic acid, not in combination. On standing, barium silicate is slowly 

 formed. 



* Freundlich, ' Kapillarchemie,' Leipzig, 1909, p. 444. 



t ' Zeit. f. anorg. Chem.,' 1903, vol. 36, p. 380. Also p. 486 of the ' Gesamnielte 

 Abhandl. von J. M. van Bemmelen,' herausgegeben von "Wo. Ostwald. Dresden, 1910. 



