PROFESSOR KOPP ON THE SPECIFIC HEAT OF SOLID BODIES. 
89 
are less applicable. But more especially can such corrections be disregarded when, as 
in the case with my determinations, other circumstances diminish more materially the 
accuracy of the results to be obtained. 
Such circumstances in my experiments are, that I worked on a small scale in every 
respect. I could only heat the solid investigated together with a liquid to 50°, and in 
many cases not even to this. In Neumann’s and in Regnault’s experiments, on the 
contrary, the solid was usually heated to near 100°, and the difference in temperature, 
T— T' (compare § IT), obtained in the latter experiments was usually thrice as great 
as in mine. In Regnault’s experiments (in Neumann’s the details are not given) the 
quantity of substance taken was, on the average, twenty times as much, and the weight 
of water in the calorimeter about eighteen times as much as in mine * : hence in the 
latter experiments the unavoidable accidental errors of observation must be greater 
than in the former. 
But there is a still more important circumstance which makes the accuracy to be hoped 
for from my experiments less than that to be expected from Regnault’s and Neumann’s 
experiments. In the latter methods the increase in the temperature of the water of the 
calorimeter is entirely, or is almost entirely produced by the solid under examination. 
In my experiments, on the contrary, this increase is produced by the glass, the solid, 
and the liquid in the glass. The thermal action due to the solid is only a part of the 
entire thermal action observed, and if from the latter that due to the liquid and to the 
glass is subtracted, all uncertainties in the assumptions as to the thermal action of the 
possible, lose their significance from necessary simplifications, and tbeir practical importance becomes finally 
very slight. The amount of correction is then to be pronounced as having no influence on the final result. 
It is more important to take into account the following. The trustworthiness of the specific heat to be 
assigned to any particular compound depends upon the certainty of the determination of the physical property, 
and upon the certainty of the knowledge of the composition of the body in question ; that is, in how far this 
compound corresponds to a given formula. The greatest trouble which can be taken in that determination, 
the consideration of all sources of error which are possible in the physical experiment, the most complete exposi- 
tion of the corrections which by developing conclusions from more or less certain assumptions may be formu- 
lated in one expression, and the most conscientious application of these corrections, — all this may be paralyzed 
by the' circumstance that the composition of the body in question is not, as it were, the ideal, not corresponding 
accurately to the formula. The partial substitution, if even to a very small extent, of one constituent by an 
isomorphous one, the attraction of water by a hygroscopic substance before and during the experiment, the 
presence of some mother-liquor in a crystallized salt, the loss of some water in drying a hydrated substance, so 
that this has not exactly the composition corresponding to the formula, — all these sources of error, which can 
scarcely be taken into account, may easily exercise an influence on the final result, whose magnitude far exceeds 
that of certain corrections applied to the physical part of the determination. It lies in the nature of the case 
that in such investigations, in some cases bodies of well-known, in other cases bodies of less well-known composi- 
tion are taken. I tried to be certain what substances could be considered as of definite composition and what of 
doubtful composition, especially where the relations between the specific heat and the atomic weight or che- 
mical composition were under discussion. 
* About sixty solids have been investigated both by Regxaglt and myself; for about thirty the weights which 
he used in his determinations are twenty times as much as in mine or more. 
0 2 
