On definite Proportions. . 451 
no longer the properties of oxygen, and can therefore no 
longer be taken into the account. The following example 
confirms the probability of this last view of the subject ; 
hut the complete explanation of this phenomenon would 
probably lead us a great step further in the doctrine of che- 
mical proportions. 
Subnitrate of the Oxide of Copper. 
T have prepared this salt in three different ways: (A) by 
gently heating the dry neutral salt, and washing away the 
undecomposed portion with boiling water ; (B) by precipi- 
tation of the neutral nitrate with lime water ; and (C) by 
precipitation with caustic ammonia, which leaves part of 
the copper in the solution. All these three methods afforded 
precisely the same salt. 
I obtained from this salt by ignition, in several experi- 
ments, 65°6 to 66 percent. of black oxide of copper, and 
the acid disengaged was in great measure liquid. Conse- 
quently this*alt contains water of crystallization, and ace 
cording to the computation hereafter to be detailed, it must 
be thus constituted ; 
Nitric acid ve ss 18°9 
Oxide of copper = .. «+. 66:0 
Water ae oa Lord 
The 66 parts of oxide of copper contain 13°2 of oxygen, 
which answer to 18*9 parts of nitric acid; if ‘considered as 
composed of nitrogen and oxygen. The remaining (5*t 
parts must have been water, and have contained 13°32 
parts of oxygen. If we wished to compute this result ace 
cording to another view of the composition of the -nitri¢ 
acid, we should obtain no relation between the oxygen of 
the base and that of the water, that can be expressed by 
an integral proportion ; for 13°2 parts of oxygen would, 
according to this view of the subject, be contained in only 
15°04 parts of nitric acid, and consequently the oxygen of 
the water would exceed in quantity that of the base, and yet 
not amount totwice asmuch. Put if we would assume that 
the oxygen of the acid were twice as much as that of the 
base, the salt would contain 30:08 parts of nitric acid, and 
3°42 of water, and the oxygen of the water would again 
observe no integral proportion io that of the base. And if 
we consider the salt as a double combination of subnitrate 
of the oxide of copper with hydrate of the same oxide, we 
still obtain no satisfactory explanation, 
The analysis of this salt appears therefore to confirm the 
idea before mentioned, that im the salts in which the nitrie 
Ff 2 acid 
