CADY AND MCFARLAND.| Composition of Natural Gas. 281 
some producing areas than for others. If the area is of such a 
character that the gas is really from one large source, then the 
gases would mix until they were of a uniform composition 
throughout. If this is the case, the well-pressure when closed 
down should be the same on the whole area, and should gradu- 
ally decrease in all wells—those which are not being drawn 
upon as well as those which are producing. If the wells do not 
show these characteristics, then they must draw their supplies. 
from separate and isolated sources, and under these circum- 
stances the gas from one may resemble that from another very 
closely or it may differ widely. The gas-producing areas in 
Kansas are very evidently isolated from one another, and the 
gases which they produce differ radically in their composition. 
A good example of this is observed in the gases from Arkansas 
City and from Dexter, Kan. These are apparently both from 
the same geological horizon and the fields are only some twenty 
miles apart, but the gases could hardly differ more from one 
another than they do if they came from the opposite sides of 
the earth. We suspect that the same sort of difference might 
be found in some of the other gas-fields of the United States if 
sufficiently extended explorations were taken up. As a rule, 
the analyses which we have recorded from other gas-fields 
have been from rather small areas, usually near the center of 
the producing field. The advantage of a more extended series 
of analyses, covering all parts of the producing field, is readily 
geen. 7 
Many of the reported variations in composition are more 
apparent than real. This is largely due to errors in interpre- 
tation of analytical data. 
As stated above, many of the published analyses are of 
doubtful value. As such we would class those showing any 
great amount of hydrogen, especially where the latter is cal- 
culated from the explosion data and the gases are measured 
Over water, as is usually the case when the more modern forms 
of gas-analytical apparatus are used. Many of the older an- 
alyses, made by Bunsen’s methods, such as those by St. Claire 
Deville, Fouqué and others, which show in a few eases the 
presence of hydrogen, are worthy of considerable confidence. 
But the modern explosion pipettes differ materially from the 
long, narrow eudiometers of Bunsen, and the temperature 
reached during the explosion is undoubtedly much higher in 
the former apparatus than in the latter. The natural result is 
