. Freezing-point Depressions for Electrolytes. 515 



should probably have greater curvature. Jones's curve for 

 moderately dilute solutions runs a little to the left of it, and 

 at higher dilutions diverges markedly to the right (as sepa- 

 rately shown). Abegg''s observations are on both sides of it, 

 but at higher dilutions his curve also goes to the right, as 

 shown. Raoult's touches it, but goes off to the right. Ar- 

 rhenius's (not shown) is considerably to the right throughout, 

 and goes widely rightward at greater dilutions. Ponsot's is 

 a little to the left. As the LB-B curve is plotted with 

 Barnes's coefficients, it is probably too low. If it be raised 

 about 2'5 per cent., the amount by which Whetham's KC1 

 coefficients are higher than Barnes's, it will lie along the 

 1-2 (1'85) line, or a little above or below that line, with its 

 upper end so directed as to run out probably at a point nearer 

 the 1-2 (1"85) intersection than either the 1-2 (1*83) or the 

 1-2 (1'87) intersection. The defective data as to ionization 

 prevent our drawing a more definite conclusion than that the 

 association indicated, if any, is less than in the case of KC1, 

 and that the depression- constant is 1*85 with a limit of error 

 of perhaps 0*02. 



The HC1 curve is interesting as exhibiting a point of 

 minimum equivalent depression. The observations on which 

 the LB-B curve is based are in good agreement. Jones's 

 curve almost coincides with it in the lower part, but goes off 

 to the right in the upper part and at higher dilutions, as 

 shown separately. Loomis's curve at higher dilutions, also 

 separately shown, goes off to the left, but in a less marked 

 manner. As drawn, the upper part of the mean curve lies 

 between the 1-2 and 2-4 (1'85) lines, and it is running out 

 to a point a little beyond the 1-2 (1*86) intersection (see 

 inset). But as it is plotted with Barnes's coefficients it is 

 perhaps too low. If raised from to 2 per cent, it might 

 run out at any point between the 1-2 (1'84) and 1-2 (1"86) 

 intersections. The data are of course very defective, but 

 they are consistent with a depression-constant of about 1*85, 

 and they seem to indicate a greater extent of association than 

 in the case of KC1. 



The L— D and J-D curves for NH 4 C1 (see inset) are not in 

 agreement, having the usual relative position of L and J 

 curves. A mean curve based on their lower parts would be 

 slightly to the left of the 1-2 (1"85) line, and directed to 

 a point considerably to the right of the 1-2 (1'86) inter- 

 section. It might thus indicate anything between a high 

 value of the depression-constant, accompanied by very con- 

 siderable association of molecules, and a constant of about 

 1*85 without association in dilute solutions, and with only a 

 slowly increasing association in stronger solutions. 



