150 



It is a well known fact that zinc cliloride in solution liydrolyzes to a 

 considerable extent so that zinc hydroxide or basic zinc chloride precipi- 

 tates, often in large quantities. The addition of a strong acid redissolves 

 the precipitate, as does also the addition of aluminium sulphate. This 

 solvent action of aluminium sulphate is, witliout doubt, due to the fact 

 that it also hydrolyzes readily in solution, the resulting sulphuric acid 

 thereby acting to repress the hydrolysis of the zinc chloride. The concen- 

 trations of tlie two salts used in the new process are zinc chloride 3%, 

 aluminium sulphate 1.5%. Sucli a solution is usually clear, so far as any 

 visible precipitate is concerned. Aluminium sulphate is thus seen to be of 

 use in preventing the precipitation and consequent loss of zinc chloride in 

 the impregnating solution, but there remains unexplained the reason why 

 two clear solutions, containing tlie same concentration of zinc, should pen- 

 etrate the wood substance to different depths, the solution of the two salts 

 apparently always going farther than the solution of the single salt. It 

 was in order to find, if possible, an explanation of this difference in pene- 

 trating power, that the experiments later described were undertaken. 



The colloidal character of both zinc hydroxide and aluminium hydrox- 

 ide is well known. It seemed likely at the outset that even in clear solu- 

 tions of zini- or aluminium salts there nnist exist i»roducts of liydnplysis in 

 a colloidal condition, consisting of sols of basic salts or even of the hydrox- 

 ides themselves. If this were the case, since the hydrolysis 

 of zinc chloride is repressed by tlu' jtrcsence of the sulpliuric acid 

 formed by the hydrolysis of aliuninium sulphate, the clear solution of the 

 composite solutioi! should t-ontain less colloidal and more crystalloidal zinc 

 salt than the siugh' solution. Tlie cellulose ■•omposing the wood cells is a 

 vegetable colloid and it is not possible for other colloids to diffuse through 

 it. In other words, the wood cells will ;ict as mcmbi'aues in carrying on a 

 l)rocess of dialysis, retaining coIloidMl material and allowing crystalloidal 

 material to pass. If the solutions pt>ssess the difference in character as 

 above described it is easy to see that a difference in penetrating power 

 nuist follow. The following experiments were undertaken to deti'rmine 

 whether zinc chloride solutions in water contain more crystalloidal zinc 

 when aluminium suljihate is ])resent than when it is not present. 



Experimental. 

 Two solutions were jircpared, each containing exactly ?>% of zinc chlor- 

 ide. To one of these aluminum sulphate was added to the amount of 1.5%. 



