1897.] PETROLEUM AND NATURAL GAS. 95 



of Taman and the plains of Roumania, concluded that mud vol- 

 canoes produced petroleum and other forms of bitumen by convert- 

 ing marsh-gas into more condensed hydrocarbons. This deriva- 

 tion of liquid and even solid bitumens like ozokerite from marsh- 

 gas as an original source was also advanced as a theory by Grabow- 

 ski, who has made special studies on Galician ozokerite. 



The simplest of the emanation theories, however, is that of the 

 Russian geologist Sokoloff, who believes that petroleum is a cosmic 

 product, formed in the crust of the earth as bitumens are formed in 

 meteorites and comets by direct union of the elements hydrogen 

 and carbon. According to this theory, the liquid and solid bitu- 

 mens represent successive stages in the condensation and oxidation 

 of simpler gaseous hydrocarbons. 



These emanation hypotheses do not find much acceptance at 

 present. The connection between the petroleum occurrence and 

 volcanic activity or hot springs seems to be far from general, and 

 may indeed be classed as local and fortuitous; the oil does not 

 issue from the earth at any higher temperature than that of the sur- 

 face, as it might be expected to if connected with deep-seated vol- 

 canic or cosmic activity ; and lastly, the most abundant oil deposits 

 are not located in the regions where upheaval and fracture of the 

 earth's crust show most strongly. 



More interest perhaps has been awakened by the theories of inor- 

 ganic origin which involve definite chemical reactions. Foremost 

 among these was that of Berthelot, who, in 1866, advanced the 

 theory that the interior of the earth contained free alkali metals, 

 and that these, when acted upon by carbonic acid or an earthy 

 carbonate at high temperatures, would form acetylides or carbides of 

 the alkali metals which decompose with water to form hydrocarbons 

 analogous to those found in petroleum. If, then, water contain- 

 ing carbonic acid gas were to reach these metallic masses by 

 infiltration and act upon them at high heat and under pressure^ 

 both liquid and gaseous hydrocarbons would result. The produc- 

 tion of metallic carbides as a product of the electric furnace, and 

 their ready decomposition for the production of acetylene gas, now 

 carried out on a commercial scale, has added new interest to this 

 theory of Berthelot's. 



This line of hypothesis was farther developed by Byasson in 1871, 

 who obtained petroleum-like products by the action of steam and 

 carbonic acid gas upon iron and its sulphide at a high tempera- 



