296 THE REVOLUTIONS OF THE CRUST OF THE EARTH. 



canoes, as in Italy, in Asia Minor, in the western part of South America, 

 in the Sunda Islands, and the Philippines, &c, rarely in the center of con- 

 tinents. The winter seems most favorable for the production of the 

 phenomenon, which fact is explained by the contraction of the crust 

 caused by the cold. As the moon must exercise upon the liquid mass 

 of the earth an effect analogous to the tides, Perry has attentively ob- 

 served the phenomena of earthquakes -with reference to this action. 

 By the comparison of more than seven thousand observations, made at 

 the commencement of this century, this savan found that earthquakes 

 are in fact more frequent at the perigee than at the apogee ; they are 

 also more frequent when the moon passes the meridian than at any other 

 relative position of the satellite to a given place.* 



It seems, however, that, contrary to what has been assumed, earth- 

 quakes do not depend upon the state of the atmosphere. The proximity 

 of large bodies of water is not necessary for the production of this phe- 

 nomenon. Attempts have been made to demonstrate the electric nature 

 of these commotions of the crust, but without success in establishing, 

 even upon general facts, the exposition of this hypothesis.! The mag- 

 netic activity of the earth has equally no relation with earthquakes, and 

 the deviation of the magnetic needle at the place alone where the com- 

 motion is felt is a proof that this perturbation of the needle must be 

 attributed to local electric currents, which may in their turn result from 

 the mechanical action employed in the production of the earthquake. 

 All things, then, considered, the commotions of the crust upon which we 

 live depeud principally upon the chemical and calorific modifications of 

 the liquid and incandescent mass, and the nature of the solid crust. 

 They are partially influenced by the substance of the nearest planets, and 

 by the cold, which determines sometimes their frequency. The other 

 causes are quite accessory and at present hardly admit of generaliza- 

 tion. 



Another series of phenomena, very often in intimate relation with 

 earthquakes, have their source in the incessant contraction of the 

 solid envelope of the globe. As the chemical activity of the fluid and 

 ignited mass is not destroyed, its effects are combined with the effects 

 of the contraction of the crust ; hence the fractures, the crevasses in the 

 solid euvelope, followed by eruptions of lava, basalt, and trachyte. We 

 may say, then, with Humboldt, that volcanic activity is only the reac- 

 tion of the interior of a planet upon its crust. 



One is often astonished that the center of the earth can furnish so 

 much material for volcanic eruptions, but an examination of the subject 

 shows that no great contraction of the crust is required to supply all 

 these eruptions. The greatest discharges of lava contain rarely more 

 than a thousand cubic meters, and seldom as much. If we imagine this 

 substance extended over the surface of theearth,the stratum thus formed 



*Inatitute, No. 10f>7, p. 201. 



tGirard, 11. v Humboldt's iTosntos, p. 145. 



