Sadtler.J ^1" [Feb. 18, 



Cherry Tree Gas (average of two analyses). 



Carbonic acid 2.28 



Carbonic oxide 



Illuminating hydrocarbons 



Hydrogen 22.50 



Marsh-gas 60.27 



Ethyl-hydride 6.80 



Oxygen 83 



Nitrogen 7.32 



100.00 

 It ■will be seen that the first three of the gases just analyzed are very 

 similar in composition, while the gas escaping from the spring at Cherry 

 Tree differs very considerably from the others. The larger amount of car- 

 bonic acid and the presence of free oxygen and nitrogen are the chief points 

 which distinguish it. It is only natural, however, that a gas escaping from 

 fresh spring water should contain these gases, as they are the gases usually 

 dissolved in spring water. As to the other constituents of these gases, hy- 

 drogen, marsh-gas, and ethyl hydride are the most important ones. In the 

 case of two of the gases, the Burns gas and the Leechburg gas, qualitative 

 tests directly proved the presence of this last constituent. The other two 

 ingredients can also be proved to be there b}' the application of Bunsen's 

 formulas to the result of the eudiometric analj'sis. Thus, as the volume of 

 gas taken for the eudiometric analysis can contain only hj^drogen, hydro- 

 carbons of the marsh-gas series and nitrogen, we need for our formulas four 

 known values; the volume taken, the contraction after passage of the 

 spark, the carbonic acid produced by the combustion and the free nitrogen 

 in the sample of gas taken. 



In the three gases first analyzed this last constituent proved to be absent, 

 so that we had only three values, and could forna only three equations, con- 

 taining three unknown quantities. These equations were: 



« + y+ 2 = v. (1). 



y+22 = C. (2). 

 3 a; + 2y + 2^' = A. (3). 



where V equals the volume taken. C the carbonic acid formed, and A the 

 absorption or contraction, consequent upon the explosion, where also x was 

 taken as hydrogen (II), y Avas taken as marsh-gas (CH^), and z was taken 

 as ethyl-hydride (C.^ Hg). When in these equations were substituted the 

 found values of C and A, I got plus values for x, y, and g. On the other 

 hand, if I assumed two constituents only, x as hj'drogen and y as marsh- 

 gas, the y was made equal to C at once, which was obviously incorrect, 

 and would have given a minus value to x. If again, I assumed x to be 

 equal to CH^, and y equal to C.^ 11^, I got false values for x and y. One 

 other assumption only was open to me, that was to take x as hydrogen, y as 

 marsh-gas (ClIJ, and z as propyl hydride (C3 Ilg). This would have given 



