CADY AND MCFARLAND.] Composition of Natural Gas. 257 
COLLECTION OF GASES. 
Much of the work of collecting the samples of gas for an- 
alysis was done by members of the University Geological Sur- 
vey, and the cost of collection was in most cases borne by the 
Survey. 
Several strong steel and tin cylinders, of a variety of shapes 
and sizes, were adapted to the needs of the work. Stop-cocks 
were put at each end of the cylinder, fitted with caps to prevent 
leakage. The tin cylinders were shipped in galvanized iron 
carrying-cases, which protected them from injury. At the gas- 
well both ends of the cylinder were opened and one end at- 
tached by a secure connection to the source of gas. The gas 
was then allowed to flow through the cylinder, at a considerable 
pressure if possible, for several minutes to insure the complete 
driving out of all air. This was found to be quite effective 
when properly carried out. Then the stop-cock at the end of 
the cylinder opposite to that at which the gas was entering 
was closed tightly and the gas allowed to accumulate in 
the cylinder. In the tin cylinders a pressure of fifty pounds 
was often carried, while in the steel cylinders the entire well- 
pressure of several hundred pounds could be reached. When 
this maximum pressure had been attained the cylinder was 
tightly closed and shipped to the laboratory. 
When received at the laboratory, enough gas was taken from 
the cylinder to fill a large glass gasometer in which the gas 
was confined over distilled water. This constituted the portion 
for the extraction of helium. At the same time a small flask 
was filled with gas for analysis. The determination of helium 
and the complete analysis were carried out immediately when- 
ever possible, and if any delay was necessitated, it was made 
as short as. possible, and precautions were taken to avoid con- 
tamination of the sample or absorption of carbon dioxid. 
METHODS OF ANALYSIS. 
For the determination of the constituents of the gas other 
than the rare elements of the argon group, the standard meth- 
ods of Hempel were used. In a 100-cubic-centimeter portion of 
the gas the determinations were made in the following order: 
Oxygen was absorbed in a phosphorus pipette; carbon dioxid 
in strong potassium hydroxid solution; ethylene and the olefins 
in fuming sulfuric acid, followed by potassium hydroxid; and, 
finally, carbon monoxid in ammoniacal cuprous chlorid. 
—15 
