I907.] AND MOUNTAIN FORMATION. 409 



of mountain forming movements in the interior of the continents 

 shows that all these effects depend upon the sea, or upon surface 

 water generally, and not at all upon the progress of the secular cool- 

 ing of the globe. 



7. Researches into the physics of the earth based on the 

 hypothesis of gravitational instability may be interesting mathe- 

 matically, but a study of the earth's surface shows that all the 

 important crustal movements now going on result from the influ- 

 ence of the sea; and hence the conclusions deduced from the 

 theory of gravitational instability are invalid, because, ingenious 

 and correct as they are on the hypothesis, they rest on a false 

 premise. 



8. In the same way it is shown that no sensible shrinkage re- 

 sults from secular cooling; on the contrary the globe seems to be 

 undergoing a gradual secular expansion, owing to the influence 

 of the sea in uplifting mountains, plateaus and other masses of 

 land, by the injection of lava saturated with steam, which forms 

 underlying masses of pumice of various degrees of density. The 

 expansion of the globe seems to be at least ten times more rapid 

 than the contraction due to secular cooling; and the great distance 

 to which the sea has withdrawn from some of the mountain ranges, 

 and the great height to which ■ the plateaus have been uplifted, 

 exhibits the mighty effects of this cause in the course of geo- 

 logical ages. 



9. The seat of terrestrial magnetism seems to be the earth's 

 crust and the surrounding atmosphere; irregularities in the mag- 

 netic forces arise from the sun and moon, while similar disturb- 

 ances accompany earthquakes and volcanic outbreaks, which sud- 

 denly agitate the air and change the electric condition of matter 

 in the field of the earth's magnetism. The irregularities near 

 islands and land masses noticed in ocean surveys indicate that the 

 permanent magnetism is confined chiefly to the solid crust, and 

 does not depend on the great incandescent nucleus of the globe. 

 Magnetic storms and aurorae depend on atmospheric charges de- 

 rived principally from the sun and moon. 



10. It is thus proved not only that earthquakes, volcanoes, 

 mountain formation, the formation of islands and plateaus, seismic 

 sea waves, and the feeble attraction of mountains long noticed in 



