On definite Proportions. 339 



Stallized in fine j^rains, having been lately prepared by the 

 solution oF iron filings, and well washed from the adherent 

 liquid : I filled with it a small glass retort H uich long 

 and i- of an inch in diameter, and exposed tt m a sand 

 bath to a heat which was gradually raised to li-niiion. . he 

 water of crystallization was collected m a receiver luted to 

 the retort, and the gas was carried off by a tuoe whicti 

 opened under water. When the mass had been weaKly 

 ignited for about an hour and a half, the evolution ot 

 sulphurous acid was almost entirely discontinued, and 

 drops of sulphuric acid appeared in the neck ot the retort. 

 I now stopped the process, and suBered the apparatus to 

 cool. The residuum left in the retort was dissolved in 

 water, and afforded a reddish yellow solution, in which 

 caustic ammonia occasioned a red precipitate, without any 

 trace of a mixture of blue or green oxide. 1 he part which 

 remained undissolved in water was dissolved by boiling in 

 pure muriatic acid, and caustic ammonia was added to 

 both solutions. The red oxide thus obtained, weighed, alter 

 icrnition, 2-4 gr. and was not in the least magnetic, even 

 when it was rubbed into a fine powder, so '.hat it contained 

 no protoKide. This oxide supposes the presence of 2-415 

 gr. of sulphuric acid in the sulphate. 1 o the solutions m 

 water and in muriatic acid as much more ot this acid was 

 added as was necessary for saturating the excess ot am- 

 monia which had been employed ; and then by means ot 

 the salt of baryta a precipitate of 292 gr. o sulphate ot 

 barvta was obtained, containing one gr. ot sulphuric acid ; 

 that is, somewhat more than one-third, but not quite halt 

 of the whole sulphuric acid, that which was distilled over, 

 as wfl! as ihat which was converted into oxygen and sul- 

 phuric acid gas, being excluded from the account. Since 

 therefore less than three-fourths of the whole quanlily o. 

 acid is required lor the conversion of tlie piotoxule into ine 

 oxide at iis exi)ense, the protoxide must contain more than 

 22-125 of oxyen for iOO parts ot iron, and the tormcr, not 

 the latter, of' The proporlions mentioned must be the true 



one 



Iron therefore has in this respect the same relation to 

 oxyen that sulphur has: in its highest degree ot oxy- 

 rrcmzation it combines with o.ily half as much more 

 oxyuenasin its lowest: unless future exi)erimeuls should 

 show that some lower stage exists in particular combinu- 

 tions of iron, which are yet rnile known; for example, 

 iu the colouring bubstanee of the blood. This circum- 

 ° y 2 Btance 



