ALUMINIUM HYDROXIDE BY ELECTROLYTES. 7 



acetic acid which had remained in the ahiminous sohition after 

 boiling. 



'' The digested solution of alumina which has not been de- 

 prived of its acetic acid by boiling, requires about twice as much 

 sulphuric acid to coagulate it as does the boiled solution, and 

 thirty or forty times as much alkali. The coagulation is complete 

 before the acetic acid is entirely saturated. 



" One of the most characteristic properties of the digested 

 and altered acetate of alumina is its loss of the power of acting 

 as a mordant. The ordinary acetate, as is well known, forms 

 a yelloW' opaque precipitate with decoction of quercitron. That 

 which has been thoroughly digested is merely coagulated by that 

 decoction — the colour of which is but little altered, and the 

 coagulum is translucent. The same effect is produced with 

 decoctions of logwood, brazilwood, etc." 



This description shows how well the solution is suited to 

 the experimental study of the phenomena of coagulation of a 

 colloidal solution by electrolytes. The solution employed in the 

 greater part of the present study is an old one made by Prof. 

 Ikeda. On analysis it was found to contain 0.457'^ of alumina. 

 It was diluted 4.4 times, so that the solution actually employed 

 contained 0.1 per cent (or more exactly 0.104?^) of alumina. 



When the electrodes of a secondary battery of 1*20 volts 

 were inserted in the two branches of a XJ-tube filled with the 

 more concentrated solution of alumina, the liquid near the nega- 

 tive electrode became milky in about half an hour, while tliat 

 in the vicinity of the opposite electrode lost its opalescence and 

 became quite clear. Tiiis observation proves beyond doubt that 

 the suspended particles of aluminium hydroxide are positively 

 charged. The same has already been found to be true for the 



