Eudiomelric Ohftrvatiom. ti^ 



The property of nitrous gas to abforb oxigen gas was firft made ufe of ; but it waj 

 then thought fufficient to compare the diminutions produced in the air with which the 

 experiment was made, and the purity of any air was confidercd to be proportioned to 

 the diminution it fufFered. 



Inquiries were afterwards direded to afcertain the real quantity of oxigen gas that 

 combined with the nitrous gas ; in order to determine by the abforption which took place 

 in the two gafes when mixed together, the relative proportions of oxigen and azote gas 

 contained in the air of the atmofphere. 



But the nitrous gas does not always afford conftant refults, unlcfs by carefully obferving 

 the fame manipulations. Ingenhoufz has long Cnce fliewn when the proportion of the 

 oxigen gas is to be determined, there is no eftablilhed law by which the diminution at- 

 tributed to the oxigen gas can be mcafured, or that of the nitrous gas which becomes 

 concentrated with it. 



I am informed by an extra£l: from the bulletin of the Philomathic Society, that 

 M. Humboldt has endeavoured, by fome very ingenious experiments, to obviate the un- 

 certainty which prevails with regard to the differences found in the nitrous gas ; and that 

 he propofes, as a flriiS: method of determining by this gas the exaft proportion of the 

 oxigen gas, feveral corrcftions of the quantities ; but I fhall prove by experiments, upon 

 which I am ftili employed with Citizen Champy the younger, that this method is founded 

 on fuppofitions which are inadmifTible *. 



The proof by hydrogen gas, for which we are indebted to Volta, has much precifion, 

 efpecially when it is made upon oxigen gas ; but it requires a complicated apparatus, 

 and the hydrogen gai may differ according to the quantity of carbon it holds in folution, 

 which may caufe the refult to vary confiderably. Neverthelefs, this method may be con- 

 fidered as fufficiently exa£l when the different airs are merely to be compared, and 

 the fame hydrogen gas is ufed to make this comparifon. But the fame accuracy cannot be 

 had when the abfolute quantity of oxigen is to be determined. The proportions by 

 weight, of the oxigen and hydrogen that enter into the compofuion of water, are known 

 with enough precifion ; but the relative fpecific gravities of the two gafes are not well 

 enough determined ; and they alter too much, according to the difference of the hydrogen 

 gas, to admit of judging exaftly of the diminution which takes place by the combuftion 

 which fhould be attributed to the oxigen and to the hydrogen gas, and determine by that 

 means the quantity of oxigen gas that is contained in the air with which the ex- 

 periment is made. 



The liquid fulphurct of alkali ofFers a double advantage in giving at the fame time 

 the comparative flate of the different airs which are examined, and the proportion of the 

 oxigen gas they contains. For all the diminution muft here be attributed to the oxigen, 



• Thefe experiments were not finiflied when I left Cairo, and I did not bring the notes of feveral refult* } 

 £br which reafon I have been obliged to repeat them, and fliall foon csramunieate them to the v^orld. — B. 



which 



