167 
1914-15.] Vapour Densities of the Ammonium Halides. 
there is a small amount of association below 340°, and that the changes 
taking place are of the type ; — 
(NHJ)2^2NH,I:^2NH3 + 2HI. 
There must be some dissociation below 340°, but this is masked by the 
associated molecules, presumably (NH 4 l) 2 , which raise the density. Above 
340° the dissociation preponderates over the association. For ammonium 
iodide the degree of dissociation cannot be calculated, even above 340°, 
because the proportion of associated molecules is unknown. It is not 
difficult to show from the data, however, that at 300° at least 24 per cent, 
of the material is associated, and at the most 17 per cent, dissociated, while 
at 380° there may be from 0 per cent, to 59 per cent, associated, and from 
11 per cent, to 41 per cent, (at most) dissociated. 
The latent heat of vaporisation varies with the temperature faster than 
it did in the two previous cases, namely from 18’0 at 300° to 24*2 at 380°. 
Johnson’s calculated value (44*5) is again based on the erroneous assump- 
tion of complete dissociation. 
When the bulbs were opened under water, for the purpose of deter- 
mining their volumes, no permanent gas, excepting a minute bubble of 
atmospheric gases derived from the water, was found in any instance. 
The unexpected nature of the values for the bromide and iodide cannot 
therefore be explained as due to decomposition of the hydrogen halide.* 
* See Journal Am. Ghem. Soc., xxxvii, p. 38 (1915). 
Chemical Laboratory of 
Columbia University, New York. 
(Issued separately May 22, 1915.) 
