316 MORRIS LOEB 



and that the variations in speed are due to the specific in- 

 fluences of the metals concerned. For, as will be seen, the 

 accelerations are not exactly the same. I give average values 

 of c 3 , always for the reaction between 1/20 normal ferric 

 and stannous chlorides. 



C3 



Without admixture 67-8 



With 1/20 A1C1 3 149-6 



With 1/20 MnCl 2 161-2 



With 1/20 G1C1 2 159-4 



[603] With 1/20 ThCl 4 157-0 



With 1/10 Aids 288-1 



With 1/10 MnCl 2 266-4 



With 1/20 SnCl 4 225-0 



With 1/40 SnCl 4 131-0 



The last two values are anomalous; but expectedly so, as 

 this compound enters into the original reaction. 



Noyes has found that the addition of free hydrochloric 

 acid alters the nature of the reaction, so that it becomes 

 one of the second order. I find that several chlorides, nota- 

 bly zirconium chloride and oxy chloride, have a similar effect. 

 The reason must be different from Noyes's explanation, based 

 on free chlorine ions. 



Another series of experiments, with the same fundamental 

 object, sought to study the influence of analogous compounds 

 on eerie sulphate. Cerium, alone of the so-called rare-earth 

 group, lends itself to studies in oxidation and reduction, 

 although virtually no work has been done in following these 

 changes quantitatively. Of the salts of the tetravalent base, 

 only the sulphate is stable in aqueous solution, and then only 

 in the presence of much free acid. On standing, its orange 

 color fades very slowly, a marked odor of ozone indicating 

 the by-product of its reduction to the cerous state. Indeed, 

 I am doubtful whether I have ever had a solution of the 

 tetravalent salt, free from any trivalent admixture. After 

 many trials I found that a fair idea of a reduction speed might 



