124 PROFESSOR 11. B. DIXON ON THE RATE OF EXPLOSION IN GASES. 
Table XVIII.—Rate of Explosion of Electrolytic Gas (1) with Nitrogen, and 
(2) with Carbonic Oxide. (K.) 
( 1 -) 
Mixture. 
2 vols. hydrogen 1 
and 1 vol. oxygen. J 
Hg-FO 
With addition 
of 1 vol. nitrogen. 
H^ + O + N 
With addition 
of 3 vols. nitrogen. 
Hj-h 0 l-Ng 
Mean rate 
2821 
2426 
2055 
Alixturc. 
2 vols. hydrogen 1 
and 1 vol. oxygen. J 
Ho + 0 
With addition 
of 1 vol. carbonic oxide. 
H3+O + K 
With addition 
of 3 vols. carbonic oxide. 
H0+O + K3 
Mean rate 
2821 
2455 
2080 
It is true that carbonic' acid and free hydrogen are found in the tube after the 
explosion ; l3ut these are due to a secondary reaction, which need not necessarily 
affect the rate ; in other words, the carbonic acid may be formed in the heated gases 
after the wave-front has passed by. 
This similarity in the retarding action of carbonic oxide and nitrogen can be 
explained in two ways. First, the carbonic oxide may be for the most part unaffected 
in the initial change occurring in the wave-front, and be, consequently, as inert as aii 
equal volume of nitrogen. Secondly, the carbonic oxide—inert towards tlie hydrogen 
and oxygen—may react with the steam produced both in the wave-front and behind 
it. The reaction between the steam and carbonic oxide in the luave-front would 
slightly quicken the explosion by developing heat. The experiments show that 
niti'Ogen retards the wave to an extent just appreciably greater than carbonic oxide. 
To test the correctness of these views concerning the role played by carbonic oxide 
in the flame, an experiment was tried with electrolytic gas diluted (1) with oxygen 
and nitrogen, and (2) with oxygen and carbonic oxide. We have seen that 
electrolytic gas, when diluted with carbonic oxide, explodes at nearly the same 
rate as when diluted with nitrogen. Now, if the views expressed above are 
correct, we may replace some of the diluent nitrogen by oxygen, and thereby make 
the explosion slower ; but if the diluent carbonic oxide is partly replaced by oxygen, 
the explosion should be considerably faster. The rates found in the two cases agree 
with this ]3rediction ;— 
