DEPRESSIONS FOR ELECTROLYTES MACGREGOR. 231 



The H 2 SO 4 (J W) curve for high dilutions, being a Jones 

 curve, is probably too far to the right, and being plotted with 

 Whetham's coefficients, is probably too high. Wildermann's 

 curve for high dilutions runs parallel to it, considerably to the 

 left. The J L B B curve, for lower dilutions, is very nearly 

 coincident with Barnes' curve, and in its lower part with Jones's 

 and Loomis's as well. But in the upper part, Jones' curve goes 

 off markedly to the right, and Loomis's markedly to the left- 

 Wildermann's is slightly to the left at the lower end, and diver- 

 ges somewhat leftwards in the upper part. Ponsot's runs nearly 

 parallel to it, somewhat to the left, and diverging to the left at 

 higher dilutions. The J L.B B curve is thus trustworthy as to 

 form ; but being plotted with Barnes' coefficients, it is probably 

 too low. The actual curve would thus appear to cross the 2 3 

 (1.85) line not far from its starting point, bend towards the 4 G 

 line, and run down below that line, finally bending slightly 

 towards it. Its course is therefore what it would be if it started 

 as a 2 3 curve, for k= 1.85 or thereabout, changed its curvature 

 a t a somewhat early stage, and tended to be transformed slowly 

 into either a double molecule curve or a 2 2 curve, or perhaps 

 both. The diagram would therefore indicate that at extreme 



o 



dilution H 2 S0 4 exists in solution in single molecules, dissociating 

 into three ions, that at an early stage and in a somewhat marked 

 manner, either doubling of molecules sets in, or partial dissocia- 

 tion into two ions, or perhaps both, that the change increases 

 slowly and steadily as dilution diminishes, and that at a concen- 

 tration of about 0.6, if the coefficients at this concentration are 

 to be trusted, the change is increasing in rate ; also that the 

 depression constant may quite readily be about 1.85. 



The K 2 SO 4 (L J A A) curve, is based on series of observa- 

 tions which in the main are in good agreement. It very nearly 

 coincides with the Loomis and Archibald curves, and Ponsot's 

 runs down slightly to the left. In its lower part it coincides 

 with the Jones curve, but in its upper part the Jones curve, 

 which is separately represented for great dilutions, runs off to 

 the right. Abegg's curve for higher dilutions runs even more 



