260 University Geological Survey of Kansas. 
APPARATUS. 
The apparatus used is shown diagrammatically in figure 6. 
A is a condensing bulb, in which, when surrounded by liquid 
air, the greater part of the methane and other hydrocarbons 
contained in the natural gas can be liquefied. B and C are 
bulbs containing twenty grams of cocoanut charcoal each. D 
and G are U-shaped glass tubes which, when immersed in 
liquid air, hold back water- and mercury-vapors. This is by 
far the simplest and most effective method of preventing the 
diffusion of mercury-vapor into the spectral tubes used. The 
actual apparatus is so constructed that D and G may both be 
immersed in the same vacuum tube. A is an automatic Spren- 
gel air-pump, the mercury being returned to the reservoir in 
the form of small drops by the suction of a Chapman pump. 
I is the receiver for collecting helium; Ff and EF are Pliicker 
tubes. F' is a permanent part of the apparatus; FH is filled with 
the gas and sealed off for later examination with a large spec- 
trometer. 
To prepare the apparatus for an analysis, a good water- 
pump was connected to the charcoal bulbs at J, and dry air . 
drawn through the apparatus to remove all gases left from the 
previous analysis. The connection with the outer air was then 
cut off and the whole exhausted as thoroughly as possible with 
the water-pump, heating the charcoal bulbs strongly at the 
game time. When exhaustion had been carried to the limit of 
the water-pump the Sprengel pump was put into operation, 
and the pumping continued until practically all the air had 
been removed. The charcoal bulbs were allowed to cool and 
then immersed in liquid air. The vacuum quickly became so 
good that a spark would not pass through the Pliicker tubes. 
The apparatus was then ready for use. 
METHOD OF ANALYSIS. 
The gas for analysis was taken from the gasometer contain- 
ing it and the amount used was determined from the weight of 
water which displaced it, making all necessary corrections for 
temperature and pressure. The bulb A was first filled with 
water, inverted, and the water displaced by the gas. It was 
then surrounded with liquid air and put into connection with 
the gas-holder. If the gas did not contain more than fifty per 
cent. of nitrogen, from twelve to fifteen liters would quickly 
pass into the bulb. A large amount of nitrogen seems to dis- 
