1814.] On the Cause of Chemical Proportions. 361 



not to conceal that I was fully aware of the impossibility of obtain- 

 ing results which would not be liable, in consequence of subsequent 

 experiments, to very considerable corrections and alterations. But 

 I thought it better to prefer the advantage which the science would 

 derive from such an attempt, imperfect as it may be, to the vain 

 satisfaction of not being ex]X)sed to criticisms and corrections. I 

 am satisfied that most of the determinations given in this memoir 

 will be hereafter coneoted in several particulars, without including 

 tliose numbers which I have been in several cases obliged to adopt 

 from analogies which may not in every case be correct. 



I ought likewise to observe that in speaking of compound vo- 

 lumes, I had no intention of determining the real volumes of the 

 substances : for example, iti saying that sulphuric acid is S + 3 O, 

 1 do not mean to say that sulphuric acid when in the state of gas 

 contains three volumes of oxygen and one volume of sulphur con- 

 densed into the bulk of one volume. On the contrary, the elemen- 

 tary volumes in combining undergo contractions in bulk, with the 

 general laws of which we are not acquainted, though we know some 

 examples. It will constitute a dift'erent, and probably much more 

 difficult, study, to determine the specilic gravhy of each sul)stance, 

 supposing it in gas ; though I think I perceive thai we shall find 

 means of calculating with considerable precision the contraction 

 which the elementary volumes should undergo in combining. When 



6 2 



I say, for example, that the subarseniate of lead is As O + li P O, 

 the nunil)er 1^ applies to the lead of which the compound contains 

 i-^ for each volume of arsenic. But it may happen that the result 

 of calculation of the compound volumes will prove that the neutral 

 arscniate is composed of one volume of acid and two volumes of 

 oxide of lead ; in which case the subarseniate ought to contain 

 three volumes of the oxide. 



Before concluding 1 ought to say a few words about a question 

 very intimately connected with the present subject; namely, IFhat 

 is the relation of the specific gruuity of solid substances with their 

 specific gravity when in the state of gas? In casting our eyes over 

 the comparative table of these weights at the end of this memoir, 

 we see too great a discordance between them to draw the conse- 

 quence that there is a relation between them. On the other hand, 

 we sometimes meet with coincidences vvliich ought not to be neg- 

 lected ; because it is possible that when the gaseous volumes are 

 rectified, and the weights of the solid substances more accurately 

 determined, by using only bodies perfectly pure : for example, by 

 ivclghiiig the metals only in the state of greatest purity that can be 

 olitained ; for hitherto they have been more or less contaminated 

 witli carburet ; and by reducing all the results to the same tempe- 

 rature, it is possible that we may hereafter perceive coincidences 

 which escajje us at present. When we compare in the following 

 table tlte weights of sulphur, pho'-phorus, and arsenic, we find that 

 xlic firit, both solid and gaseous, has a weight very nearly approach- 



