ON THE HEAT OF COMBINATION. 71 
the other acids. The sparing solubility of the arsenious acid in water pre- 
vents an accurate examination of its thermal reactions; but on repeated 
trials I obtained 0°25 F., on combining with it the same quantity of potash 
which under similar conditions gave 0°34 with nitric acid. Although a 
considerable excess of arsenious acid was taken, as proved by the fact that 
further additions produced no new development of heat, the solution still 
exhibited an alkaline reaction. The same is also well known to be true of 
the hydrocyanic and carbonic acids. In the case of bases, such as the oxide of 
copper, whose salts have all an acid reaction, this criterion will not apply ; 
but the exceptional acids are so few, and their peculiarities so well-marked, 
that they give rise to little difficulty in the experimental investigation. 
The quantities of heat developed by different bases in combining with the 
same acid are so different, that it is unnecessary to refer particularly to the 
proofs of the second law. In this case, neutralizing power has no apparent 
influence on the results, as oxide of silver, which forms salts neutral to test 
paper with the strongest acids, is one of the feeblest bases if measured by its 
thermal power. It developes, in fact, little more than one-third of the heat 
which potash does in combining with the acids. 
The more recent experiments of Graham and of Fabre and Silbermann, 
confirm the accuracy of the facts from which the second and third laws were 
deduced, that no heat is developed on mixing solutions of neutral salts or of 
a neutral salt and acid*. It is difficult however to obtain, as Graham has 
remarked, positive proof of the occurrence of combination, when such solu- 
tions are brought into contact. Fabre and Silbermann indeed are of opinion 
that acid salts cannot exist in the state of solution. 
Double Decompositions——When solutions of two neutral salts are mixed 
and a precipitate formed from their mutual decomposition, there is always a 
disengagement of heat, which, though not considerable, is perfectly definite 
in amount. It does not altogether arise from the components of the pre- 
cipitate having changed from the fluid to the solid state—as it is not always 
the same for the same precipitate—but it is chiefly connected with the latent 
heat of the precipitate. If the latter contains water of crystallization, the 
heat given out is much greater than when an anhydrous precipitate is formed... 
Experiments of this kind appears at first view to be extremely simple, but it 
is often difficult to obtain exact results, from the length of time during which 
the heat continues to be disengaged, even when the combination is aided by 
brisk agitation. 
The precipitation of the salts of barytes and lead by a soluble sulphate 
appeared to present favourable conditions for investigation, and accordingly 
I made an extensive set of experiments with these classes of salts. This is 
indeed the only part of the inquiry which I have been able to complete. A 
few other examples of double decomposition will however be noticed. 
Chloride of Barium and Sulphate of Magnesia.—Of chloride of barium 
carefully purified and dried immediately before the experiment at a low red 
heat, 16°94 grms. were taken in each experiment, equivalent to 19:00 grms. 
‘sulphate of barytes. The weight of sulphate of magnesia (dry) was 10°3 grms., 
which is a little more than sufficient to decompose completely the chloride of 
barium. The entire weight of the water employed to dissolve the salts was 
234 grms., of which one-third was taken to dissolve the sulphate of mag- 
hesia, and two-thirds to dissolve the chloride of barium. The solutions were 
contained in vessels of thin copper, the smaller of which, when filled with its 
* Slight changes of temperature may however occasionally be detected ; but in some cases 
a development, in others, an absorption of heat occurs. These thermal effects evidently arise 
from causes altogether distinct from those which produce the combination of acids and bases. 
