B. TERRESTRIAL PHYSICS AKD METEOROLOGY. 75 



must lie in the upper strata of the earth's crust ; and the 

 fluidity of the lava ejected from volcanoes is not a remnant 

 of the ancient incandescence of the earth, but is due to a lo- 

 cal evolution of heat caused by those shrinkings which have 

 always been produced by the sea and its action upon solid 

 rocks. The internal nucleus of the earth can lose but little 

 heat outward by reason of the bad conductivity of rock, 

 while, on the other hand, in the lapse of ages it must receive 

 and propagate uniformly all the heat due to the shrinkages; 

 thus a constant elevated temperature must prevail in the in- 

 terior, and therefore that increase of heat which is met with 

 every where in the earth is the result of all preceding heat 

 actions uniformly distributed by conduction in the internal 

 nucleus of the earth. 12 A, XII., 546. 



ON THE RELATIONS BETWEEN TERRESTRIAL MAGNETISM AND 



THE AURORA. 



It has been suggested by Lemstrom that terrestrial mag- 

 netism proper, although it may exercise a direct action upon 

 those discharges of atmospheric electricity which constitute 

 the aurora, yet can not contribute to the production of the 

 aurora, since that must depend upon certain conditions of 

 the different strata of air. The apex of the auroral bow is 

 rarely in the direction of the declination of the needle, and 

 the variations are too great to be explained as accidental 

 perturbations in the earth's magnetism. The formation of the 

 auroral corona is generally supposed to be an effect of per- 

 spective ; but if it were due to this alone, the auroral rays 

 should appear to unite at a very sharp angle, which is by no 

 means the case ; and Lemstrom is inclined to believe that 

 one influence of terrestrial magnetism is to bend the auroral 

 beams. In reference to this point. Professor De la Rive 

 states that although the formation of the auroral corona de- 

 pends upon the directing influence of the magnet upon the 

 electric currents which form the auroral beams, and is not 

 a simple effect of perspective, it must also depend upon the 

 direction of the passage of the electric discharges through 

 the atmosphere; a direction which itself changes wnth a con- 

 ductibility, more or less variable, of the different atmospheric 

 strata. Thus the united effect of perspective and flexure 

 ought to give to the rays a curvature and a position which 



