1897.] GENESIS OF NATURAL GAS AND PETROLEUM. 117 



Of the hypotheses proposed many have been based upon pheno- 

 mena which are more or less local when geologically considered, 

 even if occurring in terranes of wide extent. One hypothesis 

 attributes to petroleum and natural gas an origin almost cosmical. 



This hypothesis, suggested by Berthelot and afterwards developed 

 by Mendeleeff {Principles of Chemistry, Vol. i, p. 364), and restated 

 by this author in 1889, supposes that metallic carbides have been 

 produced deep in or below the earth's crust, and that these carbides 

 have been decomposed by steam giving rise to the various hydro- 

 carbons of oil and gas. Mendeleeff supposes that carbides of the 

 heavier metals, and among these especially iron, have been mainly 

 instrumental in the process. The correctness of this hypothesis, 

 which depends upon so direct an appeal to chemical facts, must be 

 tested by a consideration of the laws of chemistry in so far as they 

 bear upon the question. 



According to the experiments of Moissan (^Compts Rendus, Vol. 

 122, p. 1462), a few only of the metals are capable of forming 

 definite carbides, even at the temperature of the electric arc. 

 These are chiefly the alkali, alkaline earth and earth metals. Alumi- 

 num and beryllium are the only metals whose carbides yield a hydro- 

 carbon of the paraffin series alone (methane) on decomposition by 

 water. The action of water upon the carbides of the metals of the 

 alkalies and alkaline earths produces acetylenes. In the case of 

 the carbides of some heavier metals methane is produced in admix- 

 ture with free hydrogen and ethylene. The results of this author's 

 experiments would seem to lead to the conclusion that the carbides 

 of the earth metals only can be assumed to have participated in 

 the process of petroleum and gas formation, in accordance with 

 Mendeleeff's hypothesis, if the chemical composition of natural gas 

 as found in western Pennsylvania is taken into consideration. 



There are few elements known to chemistry whose relationships 

 towards carbon at high temperatures are better known than iron. 

 The action of steam upon iron in its pure state and when in com- 

 bination with carbon is also sufficiently well understood to justify a 

 criticism of the hypothesis upon chemical grounds. It is a fact of 

 importance that the product of the action of superheated steam upon 

 cast iron consists mainly of free hydrogen with small quantities of 

 hydrocarbons, including olefins, paraffins and others of unsaturated 

 character. 



It may be assumed, but hypothetically, that iron exists in the rocks 



PROC. AMER. PHILOS. SOC. XXXVI. 154. I. PRINTED MAY 20, 1897. 



