1897.] GENESIS OF NATURAL GAS AND PETROLEUM. 119 



Strata must prove of importance in the discussion of the subject. 

 The hypothesis of Mendeleeff would suggest that if free hydrogen 

 occurs among the hydrocarbons contained in any geological forma- 

 tion it must be looked for in those strata which are nearest above 

 the Archaean rocks, and where protection against loss by diffusion 

 upward is as nearly as possible assured by great thickness of com- 

 pact overlying beds. 



Believing the composition of natural gas from formations of con- 

 siderable depth to be a matter of interest, some tests were made 

 during August, 1896, of natural gas from a well drilled down 

 through the Trenton limestone at Stevensville, countj of Welland, 

 Ontario, Canada. This well is twenty-nine hundred feet deep and 

 stratified formations below its bottom are locally of slight depth, 

 so that, according to Mr. E. Coste, the engineer for the gas com- 

 pany, the drill has in the case of this well penetrated to within a 

 short distance only of the Archaean rocks. Shales sixteen hundred 

 feet in thickness shut off possible communication between the 

 Trenton limestone and the upper gas-producing rocks (the Medina 

 sandstone, Clinton limestone and Niagara limestone), and there 

 seemed every reason to suppose that the gas was derived exclusively 

 from very deep-lying measures. The tests were made at the well, 

 and thus the possibility of errors due to leakage during transporta- 

 tion of a sample were avoided. The method employed I have 

 described in the American Chemical Journal {ox 1894, page 258. 



Tests were also made at the well by methods which have been 

 devised for such purposes, and which have been described in the 

 same volume of the journal named, for acetylene and carbon mon- 

 oxide. The results of all these trials were negative. Numerous 

 tests have been made of gas from wells scattered over various parts 

 of western Pennsylvania which seem to justify the conclusion that 

 free hydrogen, acetylene and carbon monoxide are not found in the 

 natural gas of the region. 



The absence of free hydrogen in natural gas might be explained 

 upon the assumption that although originally present, it has, by reason 

 of its extreme lightness and ready diffusibility passed out through 

 overlying rock strata and made its way to the upper regions of the 

 atmosphere. In such case we must suppose that as a result of the 

 production of free hydrogen in the interior of the earth, the atmos- 

 phere now contains in its more rarified portion a considerable and 

 gradually increasing volume of this very light gas. 



