CADY AND MCFARLAND.|] Composition of Natural Gas. 277 
has not. The apparent heat value in these cases is of course 
higher than the corrected value would be. 
The instrument used in the determinations here presented 
was a Sargent automatic gas-calorimeter, which is so arranged 
that one operator can run a complete test on a gas without as- 
sistance. The gas was measured in a wet meter registering 
thousandths of a cubic foot and having a capacity of one-tenth 
cubic foot per revolution. Each complete revolution of the 
needle caused the closing of an electric circuit which auto- 
matically switched the flow of water from the calorimeter 
from one weighing vessel to another. At each change the 
weight of water was determined and readings made on the 
thermometers at the inlet and outlet of the water. 
During each series of measurements made on the gas, a 
sample of the gas was taken for analysis. The results given 
are averages for the heat value and for the analytical results. 
The determination of heat values was undertaken too late 
to permit any extended series of such measurements to be 
made. The difficulty of transportation of the apparatus and of 
securing sufficiently adequate places to work at the wells made 
it necessary that the work be done in the University chemical 
laboratory. It was found, also, that a supply of gas adequate 
for an extended run on a gas could hardly be carried in one 
collecting cylinder, so that determinations were limited to 
gases that could be gotten at the laboratory in goodly quantity. — 
This was found in the gas supplied to the city of Lawrence 
by the pipe-line of the Kansas Natural Gas Company. Coming 
as it does from a large number of wells scattered over a con- 
siderable area in the southern Kansas field, this gas is a com- 
posite sample that represents a whole field rather than any 
single well or set of wells. It is subject to some variation in 
composition, as will be seen from the several analyses that 
have been made upon it from time to time ever since it was 
first turned into the pipes. This variation, however, has not 
been excessive, considering the wide area from which the gas 
has come, and the results obtained may be taken as somewhat 
representative of gas of this approximate composition. It will 
not, however, be at all representative of the heat value of gas 
from the entire state. In view of the wide variations in compo- 
sition of the Kansas gases, as shown in the analyses, table 27, 
p. 270, this would be impossible, and we hope that no one will 
get such an impression. 
