tl9 Eudiametrtc Obfervatkns. 



inconvenient ; that It gives the moft pofitive refults with regard to comparing the airs, and 

 requires only one corre£llon of a fortieth part of the volume of the remaining gas, when 

 the proportion of the oxigen gas is to be determined with the fame accuracy as may be 

 obtained by the fulphuret, which appears to me to be the mofl: exaft approximation that 

 can be procured by any means known *• 



Several experiments made by Citizen Champy and myfelf in the laboratory of the 

 Inftitute of Egypt, with the fulphuret of alkali and phofphorus, in which we applied the 

 correftions fuited to the changes of temperature, as well as for the expanfion of the azote 

 by the phofphorus, have proved, that the proportion of oxigen gas in atmofpheric air in that 

 country was at leaft twenty two parts out of an hundred. We have not found in a great 

 number of experiments any difference exceeding one hundredth part; and it is more 

 natural to attribute this fnaall variation to the inevitable imperfe£lions attendant on all 

 chemical procefles, than to any real change in the flate of the atmofphere. 



Multiplied experiments I have made at Paris by the fame methods, appear to me to 

 prove, that the proportion of oxigen is there very nearly the fame as in Egypt. I muft, 

 however, remark, that I have not fo accurate a recolIeQion of the refults, as to affirm a 

 perfeft equality. It is, therefore, neceffary to repeat thefe experiments at Paris, in order 

 to eftabllfh the comparlfon beyond all doubt f. 



Several celebrated chemifts and natural philofophers give the oxigen gas a confiderably 

 higher proportion than that which I have determined ; they pretend that they have found 

 confiderable variations in the air of different places and at different times. Humboldt has 

 recently made the proportions of oxigen gas to vary from twenty-three to twenty-nine 

 centimes. But I have not perceived thefe variations at the diftance of Cairo from Paris, 



• Mr. Humboldt pretends, that he has formed a ternary combination of phofphurus a/ote and oxigen 

 (Ann. de Chim. 30 Thermidor, An. 6.) His opinion is founded on fome experiments, in which he has 

 obferved that the phofphorus produces unequal diminutions in the atmofpheric air, expofed to its aflion in 

 tubes of equal fize, and under the fame circumftances. The difference between the two tubes was from 115 

 to 156. It is on this alone that he eftabli.lies theexiftence of oxided phofphuret of azote. I have feen no 

 appearance of this difference in the numerous experiments I have made at Paris and at Cairo at two different 

 periods. 



f The experiments made at Paris gave about one two-hundreth part more of rcfidue. They coincide 

 perfeflly with thofe which I had formerly made. It appears to me likely that this fmall difference arifes 

 from the air at Cairo being very dry. At the time when I operated, the air, probably, became faturated 

 with water, and acquired by that means a flight dilation rather than a difference between the conftituent 

 parts of the two atmofpheres. But the refults prove that the proportion of oxigen gas might be determined 

 at a little above twenty-two parts out of an hundred. 



Experience will eafily decids between the opinion of Mr. Humboldt and mine. It is only required to 

 examine, whether it be true that phofphorus and the liquid fulphuret produce a uniform and conftant 

 refult, as I affirm. I do not fptak of the vivid combuftion of phofphorus, with regard to which there are 

 other confiderations to be made. When the experiment of the flow a6tion of the phofphorus is made, the 

 cylinder of phofphorus ought to occupy a large part of the height of the tube in which the experiment 

 is made, ii» order that all the atmofpheric air may be ftjbjefted to iu aftion. 



4 at 



