560 LECTURE LVII. 



haps be deduced from the cooperation of electricity, already existing in a 

 neighbouring cloud. 



It is doubtful whether the light of the aurora borealis may not be of an 

 electrical nature : the phenomenon is certainly connected with the general 

 cause of magnetism ; the primitive beams of light are supposed to be at an 

 elevation of at least 50 or 100 miles above the earth, and every where in a 

 direction parallel to that of the dipping-needle ; but perhaps, although the 

 substance is magnetical, the illumination, which renders it visible, may still 

 be derived from the passage of electricity, at too great a distance to be dis- 

 covered by any other test. 



Earthquakes* and volcanos appear to originate in chemical changes, 

 which take place within the substance of the earth : they have probably 

 little further connexion with electricity, than as causes which occasionally 

 destroy the electrical equilibrium ; for although some authors have inferred, 

 from the great velocity with which the shock of an earthquake is trans- 

 mitted from place to place, that its nature must be electrical, yet others 

 have, with greater probability, attributed the rapid succession of the effects 

 to the operation of a single cause, acting at a great distance below the earth's 

 surface. There are however some circumstances, which indicate such a 

 connexion between the state of the atmosphere and the approach of an 

 earthquake, as cannot easily be explained by any hypothesis. 



The shocks of earthquakes and the eruptions of volcanos, are in all 

 probability modifications of the effects of one common cause : the same 

 countries are liable to both of them ; and where the agitation produced by 

 an earthquake extends further than there is any reason to suspect a 

 subterraneous commotion, it is probably propagated through the earth 

 nearly in the same manner as a noise is conveyed through the air. Volca- 

 nos are found in almost all parts of the world, but most commonly in the 

 neighbourhood of the sea ; and especially in small islands ; for instance, in 

 Italy, Sicily, Iceland, Japan, the Caribbees, the Cape Verd islands, the 

 Canaries, and the Azores : there are also numerous volcanos in Mexico and 

 Peru, especially Pichincha and Cotopaxi. The subterraneous fires, which 

 are continually kept up in an open volcano, depend perhaps in general on 

 sulfureous combinations and decompositions, like the heating of a heap of 

 wet pyrites, or the union of sulfur and iron filings : but in other cases they 

 may perhaps approach more nearly to the nature of common fires. A 

 mountain of coal has been burning in Siberia for almost a century, and 

 must probably have undermined in some degree the neighbouring country. 

 The immediate cause of an eruption appears to be very frequently an 

 admission of water from the sea, or from subterraneous reservoirs ; it has 

 often happened that boiling water has been discharged in great quantities 

 from a volcano ; and the force of steam is perhaps more adequate to the 

 production of violent explosions, than any other power in nature. The 

 consequence of such an admission of water, into an immense collection of 



^ * Bertrand, Mmoires Historiques et Physiques sur lea Tremblemens de Terre. A 

 1 Haye, 1757. Michell, Conjectures concerning the Cause of Earthquakes^ Ph. Tr. 



