A. MATHEMATICS AND ASTRONOMY. 51 



others, as follows : First, the luminous phenomena, or flames, 

 seen about the summits of the mountains of Spitzbergen, 

 are also seen in Lapland, and are of the same nature as the 

 aurora borealis. Second, some phenomena of the same kind, 

 although a little different, have been observed in other coun- 

 tries; and this proves that electric discharges of the nature 

 of the aurora can take place elsewhere than in the arctic re- 

 gions. Third, the spectroscope is the surest means of decid- 

 ing, in case of doubt, as to the nature of the phenomena. 

 Fourth, in arctic countries the electric discharge preceding 

 thunder passes through the atmosphere at a lower altitude 

 than in any other countries. Fifth, the electric currents 

 which are developed, in the earth when the aurora takes 

 place are not phenomena of induction governed by the lat- 

 ter at least not in the northern regions. If they are not 

 caused by the same current which is produced by the trans- 

 portation of electricity from the upper regions of the atmos- 

 phere toward the earth, it is necessary to seek their cause 

 in the perturbation of the terrestrial current. Sixth, in all 

 probability the currents attending the aurora should be able 

 to affect the galvanometer, provided that the apparatus which 

 collects the electricity is sufficiently large, or placed suffi- 

 ciently high in the atmosphere. Seventh, as a rule, the pos- 

 itive electricity of the aurora comes from above, downward. 

 Eighth, the corona of the aurora is only a perspective phe- 

 nomenon ; but the rays themselves have a real curvature. 

 Ninth, in the spectrum of the aurora there are in all nine 

 rays, which, in all probability, are the same as the lines giv- 

 en by the gases that compose the atmosphere. Tenth, the 

 spectra of the aurora can be classified in three different types, 

 which depend upon the character of the discharge itself. 

 Bibliotheque Universelle, L., 1874, 385. 



COMPENSATION OF CLOCKS FOR THE INFLUENCE OF BARO- 

 METRIC CHANGES. 



In the new standard sidereal clock of the Royal Observa- 

 tory at Greenwich, a peculiar arrangement has been intro- 

 duced in order to counteract the influence upon the rate of 

 the clock of the varying density of the earth's atmosphere. 

 This density, as it is well known, varies principally with the 

 changes in barometric pressure; and Professor Airy has 



