192 
PROFESSOR KOPP ON THE SPECIFIC HEAT OF SOLID BODIES. 
the mean of the determinations in § 86, is much smaller than that of R 3 0 3 (==3R0), 
which is the mean of the numbers in § 85 =33*3; the atomic heat of the carbonates 
RG0 3 =20*7, as the mean of the determinations in § 86, is much smaller than 27*1, 
the number found for As 2 0 3 , Bi 2 9 3 , Gr 2 0 3 , Fe 2 0 3 , and Sb 2 0 3 as the atomic heat of 
oxides R 2 0 3 . I put the atomic heat of carbon at T8 for G, as deduced from the deter- 
mination of the specific heat of its purest variety, diamond. 
102. In the preceding paragraphs I have discussed the elements which, from the 
determinations of their specific heat in the solid free state, have a smaller atomic heat 
than about 6-4. There remain to be discussed a few elements whose atomic heats are 
also less than those of most other elements, but can only be deduced from those of their 
compounds. 
To this category belongs hydrogen *, even if the indirect determination of its atomic 
heat in the solid state is liable to the uncertainty just discussed. The atomic heat of 
water, H 2 0, is (§ 85) =8*6, and smaller by 7 than that of suboxide of copper, Gu 2 0, 
which was found in the mean to be 15*6 ; the atomic heat of hydrogen would thus be 
-|=3 '5 less than that of the elements to which copper belongs, as regards its atomic 
heat ; hence the former would be 6'4 — 3 - 5 = 2’9. The atomic heat of chloride of ammo- 
nium, N H 4 Cl, has been found to be 2(M) (§ 84 ) ; the subtraction of the atomic heats for 
N+ Cl=6-4+6-4=12'8, leaves 7’2 as the atomic heat of 4H, and therefore T7 for that 
of H. The atomic heat of nitrate of ammonia, N 2 H 4 G 3 , is 36 - 4 (§ 88); subtracting 
therefrom as the atomic heat of N 2 +0 3 , the number 27T, which has previously been fre- 
quently mentioned as the atomic heat of oxides R 2 0 3 , we h aye 9 -3 as the atomic heat 
of 4H, that is 2‘3 for that of H. I put in the sequel the atomic heat of hydrogen at 2'3. 
That oxygen has a smaller atomic heat than 6*4, follows from the fact that the oxygen 
compounds of the metals have a considerably smaller atomic heat than the correspond- 
ing chlorides, iodides, or bromides. For instance, the atomic heat of the oxides -R0 is 
as the mean of the determinations in § 85 =11T, while that of the chlorides RC1 and 
RC1 (§ 84), is 12’8, that of the corresponding bromides 13‘9, and of the corresponding 
iodides 13’4. That of the oxides, R0 2 , as the mean of the determinations in § 85, of 
. Mn0 2 , Sn0 2 , and Ti0 2 is 13*7, while that of the chlorides RC1 2 (§ 85) is 18*5, and 
of the iodides Rl 2 =19’4. Taking the atomic heat of the other elements, which are 
contained in the following compounds, at 6-4, the atomic heat of oxygen, as deduced 
from the atomic heat of the oxides R 0 (11T in the mean), is =4*7 ; as deduced from 
the oxides B 2 0 3 (27T as the mean of the oxides of this formula previously frequently 
mentioned), it is =4-8; from the above oxides, R0 2 (13*7 in the mean), it is =3-7; it is 
found (compare § 88) from K As 0 3 (25-3) to be 4T; from Pb 3 As 2 0 8 (65-4) to be 4*2; 
from KC10 3 (24-8) to be 4-0; from KC10 4 (26‘3) to be 3*4; from K-Mn0 4 (28’3) to 
be 3 ’9. In the sequel I take the round number 4 for the atomic heat of 0. 
* L. Gmelin (Handbuch der Chemie, 4 Aufl. vol. i. pp. 216 and 222) ascribed to hydrogen the same 
capacity for heat as that of an equivalent quantity of lead or mercury (H=l, Cu=31-7, Hg=100); Schroder 
(Poggend. Ann. vol. lii. p. 279) and Cannizzaro (II Nuovo Cimento, vol. vii. p. 342) ascribed to hydrogen the 
same atomic heat as that of most other elements (H=l, Cl=35-5, €hi=63*4, Hg=200). 
