Dr. Kane on the Compounds of Ammonia. 9 



rendered it necessary to confirm his formula before they could be assumed as 

 data in an investigation like the present. 



There can be obtained but one crystallized nitrate of the peroxide of mer- 

 cury : this salt is formed in small prisms, which deliquesce, except in a very dry 

 room ; when dried between folds of blotting paper, the crystals taste metallic, but 

 not acid. These crystals are decomposed by water, but only a portion of the 

 mercury is thrown down as a pale yellow powder, whilst the liquor becomes acid. 

 If the supernatant liquor be evaporated, the excess of acid is driven off, and there 

 crystallizes, on cooling, the same salt as had been previously dissolved. 



To analyze this salt, the same method was pursued as had been employed by 

 Mitscherlich, and with exactly the same result. As the analyses were but con- 

 firmatory of his accuracy, I shall not enter into their details. The formula of 

 this crystallized pemitrate of mercury is ugo. NOj + ugo -\- 2 ho, and in num- 

 bers : 



2 atoms of oxide of mercury = 202.80 



1 of nitric acid = 54.14 



2 of water = 18.00 



274.94 

 It is well known that this salt is decomposed by water, but there still remains 

 some doubt as to the constitution of the subnitrate thus generated. From the 

 variable appearance it presents, according to the method by which it has been 

 obtained, it evidently is not of constant nature ; and it is generally stated by 

 systematic writers, that by washing it can be completely resolved into nitric acid 

 and oxide of mercury. Of this nitrous turpeth, as it has been generally termed, 

 two quantitative analyses have been recorded, of which the results follow : 



Oxide of Mercury. Nitric Acid. Reference. 



Braancamp = 88.0 12 An. Chim. 54 



Grouvelle = 88.97 11.03 An. Ch. et Phys. 19 



These results coinciding so closely, and leading immediately to the formula 

 NO5 4" 4 H^o, might appear to be conclusive, but several circumstances induced 

 me to consider a new examination necessary. Thus, all other analyses made by 

 Braancamp Vere inaccurate by four or five per cent., a result to be partly attri- 

 buted to the imperfect state of analytical chemistry at the time he wrote ; and 



VOL. XIX. C 



