392 STRUCTURE, AND MODE OF ACTION 



times approximating more to basalt and sometimes to trachyte; 

 and (as we learn from the important researches of Mitecherlieh, 

 and the analogy of artificial igneous products) chemical substances 

 may have united in definite proportions in a crystalline form : in 

 all cases we recognize that substances similar in composition have 

 arrived at the surface of the earth by very different ways ; either 

 simply upheaved, or penetrating through temporary fissures; and 

 that breaking through the older rocks (t*. e. the earlier oxidized 

 crust of the globe), they have finally issued as lava currents from 

 conical mountains having a permanent crater. To confound to- 

 gether phenomena so different, is to throw the geological study of 

 volcanos and volcanic action back into the obscurity from which, by 

 the aid of numerous comparative observations and researches, it has 

 gradually begun to emerge. 



The question has often been propounded : What is it that burns 

 in volcanos what produces the heat which melts and fuses together 

 earths and metals? Modern chemical science has essayed to answer, 

 That what burns are the earths, the metals, the alkalies themselves ; 

 viz. the metalloids of those substances. The solid and already- 

 oxidized crust of the globe separates the surrounding atmosphere, 

 with the oxygen which it contains, from the inflammable unoxidized 

 substances in the interior of our planet : when those metalloids come 

 in contact with the oxygen of the atmosphere there arises disengage- 

 ment of heat. The great and celebrated chemist who propounded 

 this explanation of volcanic phenomena soon himself relinquished it. 

 Observations made in mines and caverns in all climates, and which 

 in concert with M. Arago I have collected in a separate memoir, 

 show that, even at what may be considered a very small depth, the 

 temperature of the earth is much above the mean temperature of 

 the atmosphere at the same place. A fact so remarkable, and so 

 generally confirmed, connects itself with that which we learn from 

 volcanic phenomena. The depth at which the globe may be re- 

 garded as a molten mass has been calculated. The primitive cause 

 of this subterranean heat is, as in all planets, the process of formation 

 itself, the separation of the spherically condensing mass from a cos- 

 mical gaseous fluid, and the cooling of the terrestrial strata at 

 different depths by the loss of heat parted with by radiation. All 



