138 
Proceedings of Roycd Society of Edinburgh. [sess. 
one ten-milliontli or the one hundred-millionth of a centimetre ; and, 
in full agreement with Andrews, has pointed out that the heat 
equivalent of the work is a portion of the heat of combination 
which would he given out if the two metals were brought into 
thorough chemical combination with one another. 
Hitherto, however, few experiments and no measurements have 
been made to exhibit or determine the heat of combination of zinc 
with copper, or of other pairs of solid metals. Hot only in connec- 
tion with the theory of contact electricity in particular but generally 
in respect to chemical affinities it is important that we should have 
some knowledge in regard to this question, and at the request of 
Lord Kelvin I have carried out the following experimental investi- 
gation in the Physical Laboratory of the University of Glasgow. 
The method of procedure was to dissolve a known weight of an 
alloy, and also, under similar conditions, the same weight of a 
mixture of the elements which are present in the alloy, the propor- 
tions taken being the same as those known to be in the alloy, and 
noting the initial and final temperature in each case. 
Experiments were first made on brass and on its mixed elements 
to ascertain if there was a difference in the heat of solution in the 
two cases; but before this question could be answered with cer- 
tainty, a large number of preliminary experiments were made to 
determine the most suitable conditions for the work. The nature 
and strength of the solvent, and its quantity for a given mass of 
the metals to be treated, keeping in view the advisability of obtain- 
ing a moderate range of thermometric readings, and the necessity 
of minimising as far as possible the violence of the reaction 
between metals and solvent, had to be settled. Bromine (pure, or 
diluted with water or hydrochloric acid solution in various pro- 
portions), chlorine dissolved in hydrochloric acid solution, and 
nitric acid of different strengths, were all tried, and the nitric acid 
was found to be the most suitable. Excellent results were 
obtained from nitric acid containing 57 per cent, of pure acid ; its 
density at 15° C. was P355, and its specific heat was taken as 
*634. This strength of acid was that used throughout in the sub- 
sequent experiments on zinc and copper and on brass. Different 
kinds of apparatus were used and in different ways, but that finally 
adopted is shown in the figure. 
