96 Prof. H. E. Armstrong and Dr. T. M. Lowry. [Mar. 13, 



series, viz., ELCV^SOs. But there were indications that some simpler 

 acid was also present in small quantity. 



These results applied mainly to concentrated solutions, 85 per cent, 

 of the change taking place between the limits expressed by the formula 

 HoSCVB^O and HoSO^HoO ; and virtually no peroxidation of the 

 sulphuric acid took place below the limit expressed by the formula 

 HoSCVGrLO. Yon Baeyer and Villiger, on the other hand, had 

 dealt with a product existing in dilute solution. The different conclu- 

 sions deduced from the two sets of results were, therefore, not neces- 

 sarily discordant ; it was possible that the product examined by Baeyer 

 and Villiger had been formed by hydrolysis from the " higher " acid 

 which Low^ and West's observations had indicated was present in 

 concentrated solutions. It is also to be noted that the determination 

 of the ratio of active oxygen to sulphur is sufficient to determine the 

 composition of the acid only in the case of the acid being one of the 

 HoOo'ttSOs series ; obviously, other types of " persulphuric acid" are 

 possible. 



If we consider what must be the properties of the persulphuric acids 

 generally, it is clear that whereas the salt of a dibasic acid of the 

 formula HoS0 5 would remain neutral on withdrawal of the peroxide 

 oxygen, salts of higher acids would yield more or less sulphuric acid 

 when decomposed. As a carefully neutralised solution of Caro's acid 

 becomes acid when heated, the salt originally present in it cannot be 

 one derived directly from the acid H0SO5 — assuming that this is 

 dibasic. 



We are much indebted to Mr. A. J. Cook for having made a long 

 series of determinations which show that the ratio of the increase in 

 acidnry to active oxygen lost is 



S0 3 : = 1 : 2, 



a result which finds expression in the formula H2S2O9 but is in direc 

 opposition to the formula H0SO5 ; thus, supposing the calcium salt to 

 be used : 



2CaS0 5 = 2CaS0 4 + 2 

 CaSo0 9 + HoO = CaS0 4 + HoSO, + Oo. 



The probability that the acids in question have the composition 

 suggested is considerable, if the manner in which sulphuric acid may 

 be expected to undergo electrolysis be taken into account. Strong 

 solutions of the acid may be supposed to contain both sulphuric and 

 disulphuric acids, and it may be expected that both would "per- 

 oxidise " ; on electrolysis the former would give perdi- and the latter 

 pertetra-sulphuric acid, thus — 



