198 
PROFESSOR KOPP ON THE SPECIFIC HEAT OF SOLID BODIES. 
Atomic 
weight. 
Atomic Specific 
heat. heat. 
Calculated. Calculated. 
Specific . 
heat. 
Observed. 
Quadroxalateofpot- ] 
ass . . . . 
fG*H 8 
K0 8 + 2 HO . . 
254-1 
69-7 
0-274 
0-283 
Kp. 
) 
Bitartrate of potass 
KO e . . . . 
188-1 
49-1 
0-261 
0-257 
Kp. 
Seignette salt . 
O, h. 
NaK0 fi + 4H 9 O 
282-1 
87-6 
0-311 
0-328 
Kp. 
Bimalate of potass . 
C 8 H 10 Ga G 10 +8H 2 O. 
450 
152-6 
0-339 
0-338 
Kp. 
111. The preceding synopsis shows, for the great majority of substances contained 
in it, an adequate agreement between the observed specific heats and those calculated 
on such simple assumptions. In estimating the differences, the extent must be remem- 
bered to which various observers differ for the same substance. It must be considered 
that the present better determinations of the specific heat, even those made by the 
same experimenter, for substances where it may be expected that Neumann’s law applies, 
do not exactly agree with it, not more nearly than within or ^ of the value ; and 
that for those elements which are considered here as obeying Dulong and Petit's law, 
even greater deviations occur between the numbers found experimentally and those to be 
expected on the assumption of the universal validity of this law. (These deviations, i. e. 
the differences between the atomic heats found for these elements, are seen from § 82.) 
The extent to which the experimentally determined specific heats deviate from such a law, 
Neumann’s for instance, in bodies for which calculation takes it as applying, gives of course 
the means of judging what differences may occur between the observed and calculated 
numbers without invalidating the admissibility of the calculation attempted. And it is 
as much a matter of course that, in those bodies in which a marked deviation from 
Neumann’s law has been already mentioned (compare § 95), a greater difference is found 
in the present synopsis between calculation and observation. 
I consider the agreement between calculation and observation, as shown in the synopsis 
§ 103 to 110, as in general sufficient for a first attempt of that kind. But it need 
scarcely be mentioned that I by no means consider the calculated as more accurate than 
the observed numbers, or among several numbers consider that the most accurate which 
is nearest the calculated ; for that, the bases of calculation are much too uncertain. 
The list of atomic heats given at the commencement of § 103 is scarcely much more 
accurate than were the first tables of atomic weights; but just as the latter have expe- 
rienced conlinual improvements, and thus what was at first only an approximate agree- 
ment between the calculated and observed composition of bodies has been brought 
within considerably narrower limits, and apparent exceptions been explained, so, in like 
manner, will this be the case for ascertaining what atomic heats are to be assigned to 
the elements, and how the atomic heats of compounds may be deduced therefrom. This 
much, however, may even now be said, that while formerly for many solid substances a 
statement of the specific heat could in no way be controlled, a concealed source of error 
for the determination of this property was not indicated, and an error which materially 
altered the number for this property could not be recognized, at present, even if only 
roughly, spell a control is possible. Compare § 77. 
