of Edinburgh, Session 1881 - 82 . 
821 
with the numbers calculated,* he finds that there is satisfactory 
agreement except in the case of 2KC1, 2NH 4 C1, and J(Fe 2 Cl 6 ). 
A small amount of free hydrochloric acid would account for these 
variations ; and he admits its probable presence in the case of the 
ferric chloride solution, and believes that the ammonium chloride 
solution really contains free ammonia and hydrochloric acid. He 
cannot account for the irregularity in the case of potassium chloride, 
but it must be noted that it only makes its appearance when the 
solution experimented with is concentrated, and, in any case, the 
irregularity is small compared with that found for ammonium 
chloride solutions. 
The coefficient of absorption for carbon dioxide is also abnormal 
in the case of ammonium chloride solution. Mackenzie! has de- 
termined it for solutions of various strengths of the six chlorides, 
KC1, NaCl, NH 4 C1, BaCl 2 , SrCl 2 , and CaCl 2 . The result of 
his experiments is, that whereas, in the case of strontium chloride, 
the coefficient of absorption lies between those of barium and 
calcium chlorides, as is the case with the molecular weights, that 
of ammonium chloride solution exceeds greatly those of potassium 
and sodium chlorides. I may note, in passing, a mistake that 
Mackenzie has made in his statement of results. He says (loc. cit .) : 
Fur verschiedene Salzlbsungen ist dev Einfluss des Salzes verscliieden , 
und zwar liegt die absorption beim Chlorkalium , wie sein Molecular- 
gewicht zicischen der beim Chlornatrium und CJdor ammonium, u. s. 
w.” The absorption does really occupy that position, but not the 
molecular weight. 
Further, the results obtained by Kittler,{ in his experiments on 
the electromotive force of an element composed of two solutions and 
one metal point apparently to a similar irregularity in the behaviour 
of a solution of ammonium chloride, when compared with that of 
solutions of potassium and sodium chlorides ; for when the element 
* By the formulae — 
(1) (a) = 7-35 mm. + *1683 77 . 
(2) a = 8 '30 mm. + ‘1870 77 . 
(1) Tube method ; (2) air-bubble method ; and 77 the 
number of equivalents of salt in 100 equivalents of 
water. 
t Wiedemann’s Annalen, i. 451. 
X Ibid., xv. 391, 410. 
