CADY AND MCFARLAND.| Composition of Natural Gas. 259 
For the determination of hydrogen, a portion of the residue 
from the absorption of carbon monoxid was used. It was 
found that the data from explosions were entirely too unre- 
liable for the calculation of hydrogen, and the combustion of 
the hydrogen, mixed with air, over palladinized asbestos, was 
adopted. In order to insure the combustion of only the hy- 
drogen on heating to a constant temperature of 100°, the U- 
tube containing palladinized asbestos was surrounded by a De- 
war vacuum-jacketed tube, containing water which was kept 
at the boiling-point by a platinum coil heated by the passage of 
an electric current. Several determinations of hydrogen in 
mixtures of known composition proved the efficiency of this. 
arrangement. With the electrical heating device a constant 
temperature could be maintained, while the vacuum jacket 
around the bath prevented effectually the heating of the appa-. 
ratus outside the U-tube. 
In the great majority of cases no hydrogen was od and 
it is possible that where it seemed to be present a part of the 
effect may have been due to the presence of small quantities of 
higher hydrocarbons. 
Charitschkoff*’ says that the method of fractional combus- 
tion of hydrogen and methane, with palladium asbestos, is in- 
applicable to some natural gases because of the presence of 
higher hydrocarbons, such as isopentane, etc., which are ox- 
idized with the hydrogen. 
These statements are not in accord with the work of F. C. 
Phillips,®® who has elaborately investigated the phenomena of 
oxidation of hydrogen and the hydrocarbons in the presence of 
palladium and other catalyzing agents. 
DETERMINATION OF HELIUM. 
For the determination of helium, use was made of the prop- 
erty of cocoanut charcoal of absorbing all gases except helium, 
neon and hydrogen, when cooled with liquid air. This property 
was discovered by Dewar.'®° It is our experience that hydro- 
gen is rather freely absorbed, neon much less than hydrogen, 
and helium so slightly that all but an infinitesimal trace can be 
removed by pumping with a mercury pump. 
158. J.B . Chem. Soc., 1902 : 
508), RS uss, Phys. Chem. Soc., 1902, 34, 393-395. Jour. of Chem. Soc., A II, 
159. Am. Chem. Jour., 16, 163-187. 
160. Chem. News, vol. 90, p. 90 (1904). 
