622 
PROFESSOR GRAHAM ON THE MOTION OF GASES. 
The experiments with this gas indicate a number above 0-60 for its coefficient of 
transpiration ; 0625 is 5-8ths of oxygen, while the result of Table XXXVIII., which 
is the more valuable of the two, is 06 161. But this gas is likely to vary sensibly in 
its coefficient with different capillaries, like carbonic acid ; and both its physical and 
chemical properties oppose difficulties to obtaining a correct result. The rate of air 
following this gas is made very sensibly slower in the last table ; indeed the rate of 
the capillary appears to be permanently altered, possibly from the deposition of 
sulphur from the gas. 
[I shall add, in an Appendix to this paper, a series of observations on the transpi- 
ration of gaseous mixtures, which appear to warrant the following conclusions : — 
It appears from Tables XLII. and XLIIL, that mixtures of oxygen and nitrogen in 
all proportions maintain a rate which is sensibly the arithmetical mean rate of the 
two gases. 
From Tables XLIV. and XLV., that the rates of mixtures of oxygen and carbonic 
oxide are also uniform, but that mixtures of carbonic oxide and hydrogen deviate 
greatly from the mean, inclining always to that of the heavier gas. 
In Tables XLVI. and XLVIL, that the mixtures of oxygen and carbonic acid 
maintain the mean rate ; and that mixtures of carburetted hydrogen and hydrogen 
deviate greatly from the mean, always inclining to the rate of the slower gas. 
Tables from XLVIII. to LIII. inclusive, exhibit the transpiration of hydrogen 
mixed with various other gases, particularly nitrogen, oxygen, carbonic acid, nitrous 
oxide, and nitric oxide ; and Tables LIV. and LV. contain the results of the tran- 
spiration of mixtures of carburetted hydrogen with oxygen and with hydrogen. It 
appears that while all the other gases tried appeared to maintain their usual rates of 
transpiration in a state of mixture, those of carburetted hydrogen and hydrogen are 
greatly altered ; and when the proportion of the latter gas is not more than from 5 to 
15 per cent., its rate becomes as slow as the densest gas with which it is mixed ; the 
deviation from the mean in hydrogen mixtures being relatively greatest when one of 
the gases is present in a small proportion. With an addition of so much as 25 per 
cent, of hydrogen^ carburetted hydrogen, carbonic acid and nitrous oxide continue to be 
transpired in sensibly the same times as when pure and unmixed. Hydrogen is then 
transpired of course as slowly as the other gas with which it is mixed, although the 
time of hydrogen alone is 0'44, while that of carburetted hydrogen is 055, and of 
carbonic acid and nitrous oxide 0*75. Indeed small additions of hydrogen, such as 
5 or 10 per cent., made to carburetted hydrogen, appear to prolong the time of 
transpiration ; and what is very curious, raise the time of the mixture to the empi- 
rical number of pure carburetted hydrogen, namely, 0’5625 (Table XLVII.). A 
slight retardation of the same kind may also be perceived in the similar carbonic acid 
mixtures. The transpiration-time of equal volumes of hydrogen and carbonic acid is 
0‘7339, or very little less than that of pure carbonic acid. 
