50 STUDIES, SCIENTIFIC AND SOCIAL chap. 



the general theory. For these are not volcanic, but are 

 true continental islands, forming the outer margin of the 

 old continent of South America, though now 350 miles 

 from land ; and thus, being surrounded by water instead 

 of by much heavier land, the force of gravity is somewhat 

 reduced, water having here replaced a denser mass of 

 land. 



The Argument Jrom Volcanic Phenomena. 



We now come to the more special researches of Mr. 

 Fisher, which throw so much light on the hitherto unex- 

 plained phenomena of volcanoes. By means of some 

 recent experiments on the melting-point and specific heat 

 of rocks, made at his suggestion, he arrives at the conclu- 

 sion that the average thickness of the earth's crust on 

 lands near the sea-level is only about 18 miles. Its density 

 is estimated at 2'68, water being 1, and the density of the 

 liquid substratum at 2-96.^ With these new data it 

 appears that if the melted substratum were an inert mass 

 it would have cooled at such a rate that the crust would 

 have attained its present thickness in about eight million 

 3^ears. But geologists are almost unanimously of opinion 

 that any such period as this is absurdly too small, and that 

 to account for the phenomena presented by the known 

 series of rocks and their included organic remains, the 

 very least time that must be allowed is one hundred 

 million years. The conclusion Mr. Fisher draws from this 

 discrepancy is, that the substratum is not inert but ener- 

 getic, that is, that it is in a state of movement or circu- 

 lation, convection currents continually bringing up fresh 

 heat from below and thus preventing the crust from 

 solidifying so rapidly as if there were no such cur- 

 rents. A cause of such currents is found in the friction 

 produced by tidal action in the liquid mass, which Pro- 

 fessor George Darwin has shown to be very great, and to 

 be at a maximum in the central portions.^ 



Gravity having approximately its normal value all over 



1 For these conclusions see the Appendix to Physics of the Earth's 

 Crust. 



2 This is pointed out in a paper by Mr. Fisher of a later date than 

 his volume above referred to ; in Proc Cambridge Phil. Soc, 1892. 



