462 DR LYON PLAYFAIR AND MR J. A. WANKLYN ON A MODE OF 
perties in opposition. Whereas NO, represents Cl. (one equivalent of NO,, 
replacing just as much as one equivalent of Cl), NO, is physically equivalent to 
Cl,. The wonderful deepening in colour which peroxide of nitrogen undergoes on 
being heated, led us to suspect that molecular change might also occur. This 
expectation has been fully realised. We deduce from our experiments that 
peroxide of nitrogen exists in two states; that there are two bodies having the 
same percentage composition as peroxide of nitrogen, but which are polymeric. 
The body existing at high temperatures requires the formula NO,, that exist- 
ing at low temperatures requires the formula N,O,,. 
The peroxide of nitrogen used in this research was prepared partly by distil- 
ling nitrate of lead, and partly by distilling a mixture of nitrate of lead with 
chloro-chromate of potash. The products from several operations were purified 
together, and placed in.a single hermetically sealed tube. 
‘2166 grm. of the liquid, weighed in a hermetically sealed and thin glass bulb, 
was enclosed along with a little water and some pure carbonate of baryta in a 
glass tube of 40 or 50 C. C. capacity, and containing oxygen instead of air. On 
shaking up, the enclosed bulb was broken. The apparatus was heated for some 
time in the water-bath and opened; the liquid boiled to drive off carbonic acid, and 
the amount of baryta which had dissolved was carefully estimated as sulphate of 
baryta. °5498 grm. of BaO, SO, were obtained, or 149:26 per cent. of Barium. 
By calculation, 100 parts of NO, should require 148-91 of Barium. 
Therefore, the sample of peroxide of nitrogen was pure. 
Before proceeding to take the specific gravity of peroxide of nitrogen at differ- 
ent temperatures by the methods under consideration, it was requisite to settle 
certain points relating to the reactions of the body. 
In a preliminary determination of specific gravity, made by Dumas’ method, 
some months before the chief investigation, it was found that peroxide of nitro- 
gen was completely absorbed by water, without any evolution of NO,, at any rate 
at ordinary temperatures. In that experiment the Dumas flask filled completely 
with water, when opened under water at the end of the determination ; and after 
several days even, no appreciable amount of gas had made its appearance in 
the flask. 
At first we employed air as the diluent gas in these vapour-density determina- 
tions, but were subsequently compelled to have recourse to nitrogen ; for direct 
experiment showed that, when a measured volume of air is sealed up with water 
and peroxide of nitrogen, absorption of oxygen takes place, the ultimate volume 
measuring much less than the initial volume of air. Notwithstanding this source 
of error, which would tend to make the peroxide of nitrogen seem lighter than it 
really was, the specific gravity of that body at 8°5° C. was obtained as high 
as 2°46. 
The following determinations of the vapour-density of peroxide of nitrogen 

