On definite Proportions. 283 



cated bodies, from the proportions which she observes in 

 simpler combinations. We see here, that although in the 

 neutral salt of the protoxide the proportion o\ ti.e compo- 

 nent parts is determined by the sulphur and the iron, it 

 depends no longer on these elements in the subsalts of the 

 oxide of iron, but on the acid and the base, which unite m 

 such a proportion, that the acid lakes up a cectam portion 

 of oxygen in the neutral salt, and four times as much in 

 the su'btalt : and this irregulaiity arises from the quantity 

 of ox>gen in the oxide, which is not twice as much as in 

 the protoxide, but only once and a half. Hence the affi- 

 nity of the oxycen for the iron of the oxide forcibly intro- 

 duces a proporuon for the sulphur and the iron of the salts 

 of the oxide, totally different from that which they have 

 original Iv a tendency to observe. In the salts of the oxide 

 of "copper, on the contrary, similar proportions are ob- 

 served between the sulphur and the copper, because the 

 oxide contains twice as much oxygen as the protoxide. 

 These half steps, or half intervals, as they may be called, 

 in which a subsequent combination contains only half as 

 much more as a preceding one, will require, in future ana- 

 lyses ot complicated combinations, a greater readiness in 

 calculation, and a greater precision of conception, than if 

 each sitp corresponded with a d.-uble proportion, or at 

 Jeast with a multiple of some preceding step. 



I have already mentioned, that sub^ulphate of the oxide 



of iron is deposited by a solution of the neutral sulphate of 



Ihe protoxide, and that the salt, which then crystallizes, is 



of an emerald green, and forms oblique rhombs. This salt 



is a triple combination of the neutral sulphates of the oxide 



and the protoxide. The yellow colour of the former turns 



the blueish hue of tlie latter to a purer green. T have reason 



to consider this salt as the same which Thenard supposed 



to be a completely neutial salt of the protoxide. When 



we dissolve tlie most reiiular crystals in water, we obtain a 



solution sliiihily ercenish, from which a small addition of 



ammonia throws down a subsu!i)hale of the oxide. This 



is at its first precipitation while or green, the oxide and the 



protoxide being separated together; but after some minutes 



the latter falls no longer, and the precipitate is the yellow 



pulisalt of the oxide, e~ven in close vessels. Exposed to the 



air, the solution of the salt of the protoxide absorbs oxygeri 



very rajudiv, until the triple salt is formed; and if 



\vc boil it and cool it in succession for several d\ys to- 



•iclher, we obtain a red syrup- like salt, which a«ords a 



" ■ green, 



