246 On definite Proportions: 
true minimum, lut that there are other lower proportions, ac~ 
cording to which it may be a muliiple by 6,12,0r 18. Thus, 
for example, the arsenic acid contains once and half as much 
oxygen as the arsenious$ but the black oxide of arsenic, which 
is formed by the oxidation of the metal in the open air, contains 
only one-fourth as much oxygen as the arsenious acid; and hence 
the oxygen of this acid is truly a multiple by 4, and that of the 
arsenic acid by 6, of the oxygen in the black oxide, I have 
shown how a similar remark is applicable to the acids afforded 
by sulphur, 
The progressions hitherto discovered are expressed in even 
numbers. The only exception is the progression for the com- 
binations of oxygen with ammonium, which, if we set aside the 
possibility that hydrogen may be an oxide, will advance in the 
following irregular manner: 1, 14, 8, 44, 6, 74; that is, sup- 
posing the oxygen of ammonia to be unity. Hence we may in- 
fer that this series cannot begin from the true minimum, and that 
there must be lower stages of oxidation of ammonium, than ni- 
trogen, one of which must probably be identical with hydrogen. 
If then the oxygen of nitrogen is a multiple of the oxygen of hy- 
drogen by 6, 12, or 18, the series will become perfectly regular. 
I must also remark, that from the point at which ammonium 
changes its electro-chemical modification in forming nitrogen, 
the series proceeds with greater multipliers. 
The relations, which are found between the component parts 
of more compli aied substances, are all conformable to the laws 
which regulate the simpler combinations ; for instance, the pro- 
portions of sulphur to iron in the sulphates. 
2. 
When two oxygenized bodies are combined, their proportion 
may be most readily determined from the quantities of oxygen 
which they contain, the one being always either equal to the 
other, or to an integral multiple of it. 
To these combinations helong : 
a, Salts. In neutral salts the oxygen of the acid is a multi- 
ple by 2, 3..8 of that of the base. In the supersalts the num- 
ber may be still higher. In the sw/salts, the oxygen of the acid 
is sometimes a multiple of that of the base, sometimes equal to 
it, and not uncommonly an integral submultiple. 
b. Hydrates, or combinations of water: first, with acids. 
In these the water takes completely the place of a base ; the acid 
takes up a quantity of it which contains exactly as much oxygen 
as any other base with which it would be saturated: and this 
water is totally distinct from the water of crystallization con- 
tained by some of the acids. 
Secondly : 
