Account of a new Eudiometer. $j 



elaftic form, leaving the fluid, with which it was combined, 

 unaltered in its properties. 



The internments neceflary for ascertaining the compofition 

 of the atmofphere, by means of impregnated foliations, confift 

 fimply of a fmall graduated tube, having its capacity divided 

 into one hundred parts, and greateft at the open end ; and 

 of a veflTel for containing the fluid. 



The tube, after being filled with the air to be examined, 

 is introduced into the folution; and, that the action may be. 

 more rapid, gently moved from a perpendicular towards a 

 horizontal pofition. Under thefe circumstances the air is 

 rapidly diminifhed; and, in confequence of the dark colour 

 of the fluid, it is eafy to difcover the quantity of abforption. 

 In a few minutes the experiment is completed, and the whole 

 of the oxygen condenfed by the nitrous gas in the folution in 

 the form of nitrous acid. 



In all eudiometrical procefles with impregnated folution*, 

 the period at which the diminution is at a (tand mud be ac- 

 curately obferved ; for, fhortly after this period, the volume 

 of the refidual gas begins to be a little increased, and, after 

 fome hours, it will often fill a fpace greater by feveral of the 

 hundred parts on the fcale of the tube, than that which it 

 occupied at the maximum of abforption. 



This cireumftance. depends upon the flow decompofition 

 of the nitrous acid (formed during the experiment.) bv the 

 green oxide of iron, and the confequent production of a fmall 

 quantity of aeriform fluid (chiefly nitrous gas) *; which, 

 having no affinity for the red muriate, or fulphate, of iron 

 produced] is gradually evolved, and mingled with the refidual 

 nitrogen. 



The impregnated folution with green muriate is more rapid 

 in its operation than the folution with green fulphate. In 

 cafes when thefe falts cannot be obtained in a ftate of abfo- 

 lute purity, the common or mixed fulphate of iron mav; be 

 employed. One cubic inch of moderately ftron-g impreg- 

 nated folution is capable of abforbing five or fix cubic inches 

 ot oxygen, in common procefles; but the fame quantity muft 

 never be employed for more than one experiment. 



A number of comparative experiments, made on the con- 

 ftitution of the atmofphere at the Hotwells, Briftol, in July, 



* The decompofition of nitrous ncid, by folutions containing oxyde of 

 iron, at its minimum of oxydation, is a very complex procefs. JThe green 

 oxyde, during its conversion into red oxyde, not only decontpofes the acid, 

 but likewife acts upon the water of the folution ; and ammoniac is fome- 

 times formed, and lrnall portions of nitrous oxyde aud mtrogene evolved 

 with the nitrous gas, 



Auguft, 



