On definite Proportions. 169 
turating the superfluous acid of the supersulphate of potass with 
alumina, unless the supersulphate be in great excess. Besides, 
certain bases afford double salts with almest all acids, although 
most of the acids are incapable of forming a supersalt with either 
of these bases. ‘Thus ammonia, for example, affords double salts 
with magnesia, with the protoxide of manganese, with the oxide 
of copper, and with the protoxide of zinc; and we have great 
reason to think that these bases always observe the same rela- 
tion to each other, containing equal quantities of oxygen. I shall 
adduce some few examples of double salts of the first kind, which 
may be sufficient to illustrate the law of their formation. 
Sulphate of Ammonia and Magnesia. 
Ten grammes of this salt, finely powdered, and dried in the 
sun, were heated in a platina crucible, and then ignited. They 
afforded exactly one-third of their weight of sulphate of mag- 
_nesia, whence the magnesia is found to amount to 11:11 per cent. 
and its oxygen to 4°45 per cent. of the whole weight. I now 
mixed, with these 3°334 gr. of sulphate of magnesia, a quantity 
of sulphate of ammonia, in which the oxygen of the ammonia 
amounted to "443 gr. that is, 4°18] gr.: the two salts were dis- 
solved in boiling water, and dried on a glass dish in the sun- 
shine: their joint-weight was found to be 10°006 gr. Conse- 
quently the sulphate of magnesia had taken up 2°49 gr. of water, 
containing 2-2 gr. of oxygen, that is, five times as much as either 
of the bases contained. But since the sulphate of ammoma 
contains water of crystallization of which the oxygen is equal to 
twice that of the base, the whole water of the salt contains seven 
times as much oxygen as that of each of the bases: and the re- 
spective portions of oxygen in the different substances entering 
into the combination are as 1, 1, 6, and 7. 
Sulphate of Ammonia and the Oxide of Copper. 
Ten grammes of this salt, finely powdered, and dried in the 
sun, were mixed with lime in a small retort, and the ammonia was 
expelled in the same manner as in the analysis of the sulphate of 
ammonia. The apparatus had lost ‘$27 of its weight. Ten 
more grammes, dissolved in water, were mixed with about as 
much of the carbonate of potass, as was required for the satura- 
tion of the sulphuric acid, and then evaporated to dryness. When 
again dissolved in water, they left behind carbonate of the oxide 
of copper. The fluid, which had a slight excess of alkali, exhi- 
bited, by the test of sulphuretted hydrogen, a slight trace of cop- 
per. The oxide of copper obtained, weighed, after ignition, two 
grammes, These contain ‘3932 gr. of oxygen; and the *827 
gt. of caustic ammonia °3897 gr.; so that the two bases esi 
equa 
