1820.] Physical Science during the Year \%\d. 5 



into carbonic acid gas and water ; but sometimes when the sub- 

 stance analyzed is of a fixed nature (Kilkenny coal, for example), 

 we do not produce its entire destruction till we draw the coated 

 end of the tube into the choffer, and heat it fully red-hot. 



The capacity of the three tubes of which my apparatus con- 

 sists, supposing them empty, amounts to two cubic inches ; but 

 as the copper tube is completely filled with peroxide of copper, 

 and the glass tube with muriate of lime, the capacity is reduced 

 to 0*5 cubic inch. It is this considerable quantity of common 

 air with which my gaseous products are mixed that constitutes 

 the defect of my apparatus. T'le mean quantity of common air 

 is 0*5 cubic inch ; but sometimes it is as small as 0-44,, 

 and sometimes as great as 0*60 cubic inch. This difference 

 obviously depends tipon two things : 1, the degree of heat which 

 I apply ; 2, the care with which the tubes have been filled v.ith . 

 peroxide of copper and muriate of lime. At present, when the 

 substance subjected to analysis contains azote, and when conse- 

 quently the gas Vv'hich I obtain is a mixture of carbonic acid and 

 azotic gas, 1 cannot be absolutely certain of the bulk of the 

 azotic gas. My method is to make at least six trials, to reckon 

 the bulk of the common air 0".3 cubic inch, and to reckon the 

 bulk of residue after the carbonic acid gas has been absorbed by 

 potash, exceeding 0*5 cubic inch, to be azote. This, in such a 

 nuinber of trials, cannot deviate far from the truth ; but the 

 necessity of repeating the experiment so often as six times is a 

 serious inconvenience. When the substance analyzed contains 

 no azote, I obtain a very good result by taking the mean of two 

 trials. 1 have no doubt that I shall succeed in correcting this 

 defect by making my apparatus on a smaller scale. This I 

 intend speedily to try. 



In analyzing animal and vegetable substances by means of 

 peroxide of copper, I have observed that if vv-e make use of the 

 same peroxide of copper for a number of successive experiments,' 

 heating it to redness after each, after a certain time it becomes 

 useless as a reagent. It ceases to give out oxygen to the sub- 

 stance heated with it. Of course the results are nearly the same . 

 as if the substance subjected to analysis were merely heated in 

 contact with so much sand. To bring back the peroxide of 

 copper to its original state of goodness, I find it necessary to 

 dissolve it in an acid, and precipitate it by an alkali. This 

 change in the state of the peroxide is not owing to the oxygen 

 which it contains being diminished, but to the density of the 

 powder being greatly increased. After a certain portion of per- 

 oxide of copper has been employed in twelve successive exper:- 

 ments, its bulk is reduced to one-third of what it was at first ; 

 but its weight is not sensibly altered. 



2. I may here notice the method which I commonly employ 

 to detect the quantity of lime, wl\en that lime has been precipi- 

 tated by oxalate of ammonia. It is connected with experimental 



