500 JOSEPH BARRELL 



weight to the direct evidence regarding the nature of the earth's 

 interior. Because of the ease and certainty of laboratory studies 

 there is a tendency to treat the interior of the earth as though 

 it were incapable of speaking for itself through the evidence of 

 geophysics, geodesy, and geology, but must remain forever a play- 

 ground for the speculative imagination. Largely unknown the 

 nature of the earth's interior is and long must be; laboratory 

 studies on the influences of heat, of pressure, and of chemical 

 composition, upon the physical state of the crust, must constitute 

 the paths which guide in the search downward into the unknown; 

 but the final test of hypothesis must be the direct testimony of the 

 earth itself. 



The rectihnear projection of surface conditions is based on the 

 assumption that the temperature gradient is a straight line to 

 great depths, or that strength, or density, or porosity, as the case 

 may be, is not changed by the pressures of the interior. Such 

 assumptions lead to views more or less in opposition to those 

 reached in the present investigation. They must, therefore, be 

 discussed to some degree. An illustration of these dangers of 

 reasoning by unchecked extrapolation is suppHed by a paper 

 written by Arrhenius,^ selected for discussion because of the emi- 

 nence of the author in the fields of physics and chemistry, the 

 definiteness with which his conclusions are stated, and the wide 

 citation which this paper has achieved.^ In this paper the argu- 

 ments are given in favor of a gaseous nature of the interior of the 

 earth, carrying forward an idea first suggested by A. Ritter in a 

 series of "Researches on the Height of the Atmosphere and the 

 Constitution of Gaseous Heavenly Bodies." 



From the rate of increase of temperature with depth Arrhenius 

 argues that at a depth of 40 km. the crust must pass into a molten 

 condition, but one which, because of pressure, is a viscous and 

 highly incompressible liquid. At a depth of some 300 km. the 

 temperature, he states, must be above the critical temperature of 



' "Zur Physik des Vulkanismus," Geol. Foren. i Stockholm, Forhandl., XXII 

 (1900), 395-419- 



^ See for example its presentation by A. Geikie, Text Book of Geology, Vol. I (1903), 

 pp. 71-74. 



