tti Eudiomeliic Obfctvations. 



which is divided between the oxigen and nitrous, or the hydrogen gas In the preceding ex- 

 periments. There is no neceflity for any other correftions than thofe which the difference 

 of temperature and the preffure of the atmofphere require between the inllant that the 

 experiment is began, and that when the diminution of the air is meafured. 



It cannot be fuppofed that the abforption of the oxigen would be incomplete when 

 ■water fufEciently charged with fulphuret of alkali is ufed ; for there is a great difference 

 between the power which fulphuret exerts on the oxigen gas, and the weak adion the 

 azote in the gafeous flate may employ on the oxigen. And if any diaiinucion be found 

 in the azote which is entirely feparated by this means, by afterwards mixing it with 

 nitrous gas, I (hall fhew that this diminution ought not to be attributed to the oxigen. 



It mud be admitted, that it cannot be faid that the real volume of the azote gas is that 

 which was obtained ; becaufe azote may hold fulphuret in folution, or very probably the 

 fulphurated hydrogen which always exifts in liquid fulphuret ; and it certainly has the- 

 fmell, though this fmell difappears on wafliing it in a little water without any fenfible 

 diminution of volume, fo that the difference of volume which takes place in this folution 

 *nufl be extremely fmall. 



There can be no apprehenfion that the azote fliould be abforbed by the fulphuret, for 

 If this abforption took place it would continue. But the quantity of azote remains con- 

 ftantly the fame after the oxigen is abforbed. 



It is therefore, poffible to determine the proportion of oxigen which is found in any 

 quantity of air, by means of the liquid fulphuret, with all the preciCon that can reafonably 

 be expcfted in chemiftry. 



The inconvenience of this method is, that the fulphuret afts flowly, and requires longer 

 lime, and particularly a low temperature ; befides it gives no certain indication when the 

 abforption is compleated, except the ceffation of this diminution^ whicli requires ftill 

 more time to be afcertained. 



Guyton has propofed to ufe the dry fulphuret, by applying the heat of a candle to an 

 apparatus he has defcribcd. I have not tried this method, but it appears to me, that fo 

 fmall a portion of matter may not come into good contaft with fo large a volume of air, 

 by which means the abforption of all the oxigen may be rendered uncertain. The ex- 

 periment does not prefent any indication that proves its certainty. 



I have recommended the flow combuftion of phofphorus. I firft place a cylinder of 

 phofphorus fixed on a ftick of glafs in an upright veffel, in which the air fubjefted to 

 experiment is contained under water. If the temperature of the room be high, tKe veffel 

 is immerfed in water, in order that the phofphorus may not melt. For the evaporation of 

 the furface of the water keeps it at a temperature fome degrees below that of the at- 

 mofphere.^ Thus, during the experiments that I made here, the thermometer remained at 

 about 36° degrees of the thermometer centigrade, and the bath in which the experiments 

 were made was at fix degrees below this temperature. Immediately after the phofpherus 

 Je introduced into the air, a cloud is formed, which defcends and mixes with the water. 



Whea 



