ON EVAPORATION AND DISSOCIATION. 
87 
that combination is only partial. Marignac calculated the latent heat of vaporiza¬ 
tion of NH^Cl and found it to be 706 calories; while Favre and Silbermann give 
the heat of combination of NH g and HC1 as 743‘5 or 715'5 calories, calculating by 
different methods. Horstmann determined the vapour-pressure by distilling the 
substance in an iron tube closed by a cork in which were three holes; one for a 
manometer, one for a thermometer, and one for exhausting air, and so lowering 
pressure. And from his observations, the details of which he does not record, he con¬ 
structed a curve showing the relations of pressure and temperature (‘ Berichte,’ 1869, 
p. 137). On referring to this curve it will be seen that his results closely agree with 
those obtained by us. Than contradicts Deville’s results as regards rise of tem¬ 
perature on mixing the gases, by experiments conducted on a different principle; at 
350° he found no evolution of heat, or, more correctly, noticed no contraction, while 
between 330° and 340° contraction occurred ( £ Annalen,’ 131, p. 129). The well-known 
experiments on the diffusion of the hot vapour of ammonium chloride point to at least 
partial dissociation. 
§ 36. Vapour-pressures of Ammonium chloride in Barometer-tube. 
For temperatures between 98° and 280° an ordinary barometer-tube was used. As 
mercury-vapour at 280° exerts a pressure of 155T7 millims., it was first advisable to 
determine whether a mixture of mercury-vapour and another vapour on which it had 
no action could be relied on as giving the sum of their pressures. 
A barometer-tube was boiled out, and jacketed with bromonaphthalene, no sub¬ 
stance being introduced. At a pressure of 755'9 millims., and at a temperature as 
determined by the vapour-pressure of bromonaphthalene of 2S0'1° (see § 9), the vapour- 
pressure of mercury, after all corrections had been introduced, was found to be 15 8 '8 
millims. On cooling there was a certain amount of air present which had been given 
off from the walls of the barometer-tube. The pressure exerted by this air was read 
at 24°, and was found to be 10'6 millims. The volume occupied by this air was (23'3 
millims. X by area of the tube), while at 280d° it was (159'3 millims. X by area of 
tuba); hence the pressure at the higher temperature was found by the equation 
10-6X23-3X553T 
159"3 X 297 
= 2‘9 millims. 
The vapour-pressure of mercury is therefore 158'8 — 2’9 = 155 - 9 millims. This is identi¬ 
cal with that given by Regnault, for mercury-vapour at 280T 0 .* This tube without 
being emptied was inverted and again boiled out in vacuo. Bromonaphthalene was then 
introduced, and the tube was jacketed with bromonaphthalene vapour, boiling under 
reduced pressure. The temperature was occasionally raised by increasing the pressure, 
and readings were taken. The results obtained were always too low, and were vari¬ 
able. The cause of this turned out to be the slow rate at which the mercury-vapour 
diffused through the layer of liquid bromonaphthalene on the surface of the mercury, 
* From our revised table of tbe vapour-pressures of mercury the pressure at 280T° is 157'7 millims. 
