of Edinburgh , Session 1869 - 70 . 73 
volumes of the halogen salts in the above table, it is clear that 
the equivalent volumes increase, chloride of potassium being 37, 
bromide 44, and iodide 55' 3, and their relative stability diminishes. 
The equivalent volumes of chlorine, bromine, and iodine are iden- 
tical in the liquid state ; and thus the formation of the respective 
potassium compounds is one of the results of unequal condensa- 
tion, the co-efficient of contraction in the formation of chloride of 
potassium being 046, bromide 0*29, iodide 0*23 per unit volume. 
Their formation is attended with the evolution of very different 
amounts of heat. The following table contains some of the con- 
stants found with reference to combination and solution : — 
Constants of Group. 
Contrac- 
tion per 
Unit 
Volume. 
Total heat. 
Heat of 
solution. 
Diffusion 
times 
(relative). 
Co-efficient of 
expansion per 
equivalent 
volume. 
Specific heat 
per atom. 
KC1 
046 
97086 
3874 
74-5 
0-001429 
12-88 
KBr 
0-29 
85666 
4522 
119 
0-001848 
13*47 
KI 
0-23 
72721 
4847 
166 
0*002358 
13-60 
Generally speaking, the number found for bromide of potassium 
is nearly the mean of those attached to chloride and iodide. A 
similar observation has recently been made by M. Yalsen in exa- 
mining the equivalent capillary constants of these bodies. Look- 
ing at the atomic thermal number, there is a far greater likelihood 
of condensation taking place in the bromide and iodide of potassium 
in the combined state, than in case of chloride, seeing that it 
would be relatively far more difficult to condense. But neither the 
chlorate, bromate, nor iodate can be produced through the direct 
addition of oxygen to the respective halogen salt. And the 
chlorate, it is well known, evolves heat on giving off its oxygen, 
and thus necessitates an absorption of heat during combination. 
It is just possible that the heat produced during the decomposition 
is the result of the necessary expansion of volume in the chloride 
of potassium in combining with oxygen, and its return to its 
normal volume on losing it. It makes no change in volume to 
suppose that, in the one case, the oxygen is added as a whole to 
the chloride of potassium, or, in the other, that it is between the 
