ON EUDIOMETRY. 1Q7 



For instance, in your Chemkal Statics you seem to insl- Mistaken by 

 nuate, that Mr. de .viarty a^^cribes to sulphurcts the pro- ^t^'^JJ^^"^^* ^* 

 pertv of absorbing nitrogen from their nature; and von 

 Humboldt and Gay-Lussac in their work on Eudiometry 

 express this opinion still more affirmatively. This errour 

 arose from the abstract. Mr. de Marty says expressly in Sul>ihure!% 

 his paper, that a hot sulphuret acts as any otl.er liquid f^^nuJog^^ 

 would do, that had been deprived of the quantity of nitro- 

 gen it is naturally capable of absorbing j and thus he ac- 

 counted lor tli€ variatioiis he e?cperienced on operating with 

 greater or smaller quantities of sulphuret. 



Von Humboldt aad Gay-Lussac say too, that Mr.de Mar- Proportion d 

 ty fixes the propprtion of oxigen in the atmosphere to be atm^osr>her« * 

 between '21 and '23; and hence they draw an argument 

 against the method of operating with sulphurets, which in ^ 

 4:onsequence appears to them much less accurate than the 

 proof by hidrogeu gas. But this uncertainty between '21 

 and '23 Mr. de Marty experienced ouly in h s first experi- 

 ments : and it was from this very variation, and a wish to 

 diminish its extent, that he was led to discover the eriour 

 occasioned by the absorption of nitrogen by the sulphuret, 

 when it is deprived of this gas by heat ; so that when he had 

 completed the improvement of his method, the results were 

 constantly restricted between •21 and '22. 



This being once thoroughly ptoved, I do not see what As a test of it, 



advantage the analysis by hidrotcen eas has over that by the '^"^'^^g^" "o^ 



,1 1 . '. p , . . 1 . ,. , superior loah- 



enlphuret, when it is pertormed with due precaution, which quid swlphuret 



should always be a matter of course. It certainly has not in requiring 

 the advantage of requiring less time; for by operating as less time, 

 Mr. de Marty does, any one who hns acquired a little habit 

 of making the experiment can perform this in five minutes. 

 It has not that of greater simplicity : for the sulphu'et re- bein«rTnoTe 

 quires only a graduated tube, and a ground stopple bottle ; simple, 

 while for the hidvogen gas at least a small eudiometer and 

 an electrophorus are necessary ; and what is very inconve- 

 nient, the latter must be kept in a state to give sparks, 

 which is not very easy on mountains and when travelling, 

 particularly if the air be loaded with moisture. Lastly I 

 will add, that it has not the advantage of superior accuracy: or giving a 

 for when it is once proved, that the sulphuret will absorb "^^^^f accurate 



only 



