182 
PROFESSOR KOPP ON THE SPECIFIC HEAT OF SOLID BODIES. 
universal. The oxygen compounds of the metals correspond to it in general the less 
the greater the number of oxygen atoms they contain as compared with that of metal. 
The mean atomic heat of the oxides EG in § 85 is 11T, and the quotient ^=5*6. 
The quotient for the oxides R 2 0 3 and R 2 0 3 (even excluding the determinations of 
alumina and boracic acid) is only ?7 j?= 5*4; for the oxides R0 2 (even excluding the 
determinations for silicic acid and zircon) only ^=4*6 ; for the oxides R0 3 , the mean 
of Regnault’s determinations only ~=4 - 7. Still smaller is the quotient for com- 
pounds which contain boron in addition to oxygen ( e . g. for the compounds R B0 2 
(compare § 87) it is only — =4*2; for boracic acid, B 2 0 3 , it is only I^=3*3), and also 
for compounds which contain silicium in addition to oxygen (it is ^=3*8 for silicic 
acid, Si 0 2 , compare § 85), or which contain oxygen as well as hydrogen (for ice, II 2 0, 
it is only ^=2*9*, compare § 85), or which contain hydrogen and carbon besides 
oxygen (e. g. it is only ^=2*6 for succinic acid, 0 4 H 6 O 4 , compare § 89). It may be 
said in a few words what are the cases in which this quotient approximates to the 
atomic heat of most elements, and what the cases in which it is smaller. It is near 6 ’4 
in those compounds which only contain elements whose atomic heats, corresponding to 
Dulong and Petit’s law, are nearly = 6*4; it is smaller in compounds which contain 
elements not coming under Dulong and Petit’s law and having a much smaller atomic 
heat than 6*4, and which are recognized as exceptions to this law, either directly, if 
their specific heat has been determined for the solid condition (compare § 92), or in- 
directly, if it be determined in the manner to be subsequently described. 
94. The determinations of specific heat given in §§ 83 to 89 contain the proofs 
hitherto recognized for the law that chemically-similar bodies of analogous atomic con- 
stitution have approximately the same atomic heat ; and a considerable number of new ex- 
amples of the prevalence of this regularity are given by my determinations. The groups 
of analogous compounds need not again be collated, as Neumann has done for a smaller 
and Regnault for a larger number of groups and for individual elements contained in 
them. What I will here discuss is the prevalence, beyond the limits of our previous 
* Considering the atomic heat of liquid water to be 18, Garnier (Compt. Rendus, vol. xxxv. p. 278) 
thought that the quotient obtained by dividing the atomic weight by the number of elementary atoms in one 
atom of the compound, -U =6, came near the atomic heat of the elements. But it requires no explanation 
that, in a comparison with the atomic heats of solid elements and solid compounds, that atomic heat must he 
taken for the compound H 2 9 which is obtained from the specific heat of ice, and not from that of water. 
Garnier is not alone in his error, which is rather to he ascribed to the circumstance that formerly both solids 
and liquids were compared, as regards their specific heat, in considerations how this property is influenced 
by the composition. Hermann more especially (Nouveaux Memoires de la Societe des Naturalistes de 
Moscou, vol. iii. p. 137) compared liquid water with solid compounds, as did also Schroder (Poggendorff’s 
* Annalen,’ vol. Iii. p. 279) and L. Gmeein in an early discussion of this subject (Gehler’s ‘ Physicalische 
Worterbuch, neue Bearbeitung,’ vol. ix. p. 1942), while he subsequently (Handbuch der Chemie, 4. Aufl., vol. i. 
p. 220) more correctly compared the specific and the atomic heat of ice with that of other solid compounds. 
