CADY AND MCFARLAND.| Composition of Natural Gas. 283 
mean of the values for the different equations agreed well with 
this residual nitrogen as actually determined. Consequently, 
we believe that the figures which we have given in our table of 
results are not far from correct. 
That actual variation in the composition of the gas from a 
single well may take place seems to be shown by the determina- 
tion of nitrogen by Phillips in the gas from certain wells, the 
samples being taken from time to time for a considerable 
period.'® For example, the nitrogen content of the Shields 
well, fourteen miles west of Pittsburg, ran as follows: 
Percentage of nitrogen. 
ATUGUS DB, LOO sie: Maire oko aie eo eae is Ceres aires 125 1.26 
Ee LOTTE OM ON le Web sai ails va, vteete ss) eus oA ba; o) Syecsi ever eehelsls 2.67 2.68 
Aor Gy LSD eels a ced coere eee MeO ani es let ite Re CHE aI ere IL 1.80 
A\rorell QO), IS ae: 5 arate Gre OGIO © Oh eet. Osi Range mr rete 1.85 1.85 
HUM I WO s sas od cave ee emicae een icmeno icity ite Symes ica eee RIG) LUO 
These fluctuations are evidently much greater than the ex- 
perimental error and one must conclude that the variations are 
real. An explanation of the cause of these changes is not easy. 
Some have suggested that it is due to the gas being liquefied 
by pressure and then gradually boiling off the more volatile 
constituents. This certainly cannot be the cause, because the 
critical temperature of methane, above which temperature it 
cannot be liquefied by any pressure, however great, is —81.8° 
Centigrade, while the temperature of the gas in the earth is of 
course much higher than this. 
In case the gas comes from several strata, separated from 
each other by impervious layers, it is not unlikely that the gas 
from the various strata will differ in composition and also that 
the relative proportion of the outflowing gases from each 
stratum will vary with the water condition and the demand 
upon the well. It is the rule that the deeper the point at which 
the gas is struck the higher will be its pressure. If two gas- 
sands are passed through by the well they will probably have 
different pressures, and if the well is working to something 
like its full capacity both sands will be contributing gas; but 
if the flow is reduced until the well-pressure is as high or 
higher than that of the weaker sand, only the stronger sand 
will furnish the gas. If there is a difference in composition of 
the gas from the two sands it is very obvious that the compo- 
sition of the gas will vary with the flow. If the well is so far 
closed down that the pressure rises higher than that of the 
IMG!) 1Pixoxe, ANT, AN@RKEL, GA, 7S}, alkskets}, 
