114 ANNUAL OF SCIENTIFIC DISCOVERY. 



in the circuit a luminous arc 3 millimetres long : copper, 23 cubic cen- 

 timetres, brass 26, iron 27, coke 29, zinc 35, tin 45. 3. The differ- 

 ence in weight in the two poles, after experimenting, varies chiefly 

 with the elevation of temperature, other circumstances being equal. 

 With coke and iron, it is always the positive pole which has dimin- 

 ished the most in weight, but with all the other metals it is the nega- 

 tive. With coke and iron the positive point is found corroded, with 

 the others the negative. With poles of brass it is found that the pos- 

 itive pole always increases in weight. 4. The elevation of the tem- 

 perature of the poles, the conducting power of the arc, and the quan- 

 tity of matter destroyed, are greater in common air than in rarefied air 

 or in hydrogen gas. The quantity of matter disappearing in the pro- 

 duction of the voltaic arc varies with the position of the arc relative to 

 the magnetic meridian, being greater when the arc is perpendicular to 

 the meridian than when it is in the same plane. It is shown that the 

 matter detached from the two poles passes from the one to the other. 

 The voltaic arc, like the electric spark, decomposes gases through 

 which it is transmitted. 



M. Despretz, in a paper presented to the French Academy on April 

 1, gives the following as the results of experiments upon the voltaic 

 arc. 1. The length of the arc increases more rapidly than in propor- 

 tion to the number of elements in the battery, when it is placed end to 

 end, and the increase is more rapid for small than for large arcs. 

 Thus the arc produced by 100 elements is almost quadruple that given 

 by 50 elements. 2. If the battery is so arranged as to unite the simi- 

 lar poles, that is, " quantitatively," the length of the arc does not in- 

 crease in proportion to the number of elements. Thus the arc of 100 

 elements being 25.2 millimetres, it is only 69.2 mil. for 600 elements 

 arranged in six parallel series, although when placed end to end they 

 give an arc of 183.5 mil. 3. If the positive pole is below, the arc is 

 not so long as when it is above. 4. In a plane perpendicular to the 

 magnetic meridian, the arc is larger when the positive pole is towards 

 the east than when it is towards the west. 



ON THE CAUSE OF THE AURORA BOREALIS. 



" WITH regard to the origin of the aurora borealis, it seems natural 

 to attribute it to the electric fluid contained in the atmosphere, which 

 at great heights, where the air is rarefied, must become luminous, as 

 under the receiver of an air-pump, and in the barometer vacuum ; this 

 hypothesis would acquire a great probability if we succeeded in prov- 

 ing by direct experiments that magnetism exerts an influence upon 

 electric light." 



This extract from a memoir by Morlet on the aurora borealis in- 

 duced M. de la Rive to communicate to the Paris Academy the fol- 

 lowing experiment, showing the influence of magnetism on the light 

 produced by ordinary electrical discharges. " I introduce into a glass 

 globe, by one of the two tubulures with which it is furnished, a cylin- 

 drical iron bar, of such length that one of its extremities reaches near- 

 ly to the centre of the globe, while the other extends a short distance 



