METEOROLOGY AND ALLIED SUBJECTS. 309 



tious during this latter were among the most extensive ever taken. In 

 general the disturbances are the greatest on telegraph lines that inter- 

 Kcct the meridians; but decided exceptions to this rule are sometimes 

 i-ecorded. 



Again, in general the optical phenomena follow after the disturbances 

 experienced on the telegraph lines, and the latter die away rapidly after 

 the maximum of the optical display, but sometimes the earth currents 

 disappear with the hrst appearance of the optical phenomena. It is 

 jwobable that a thorough investigation of this subject would be advan- 

 tageous both for science and the telegraph companies. 



As to the explanation of the origin of earth currents, Kuhn in 1861 

 believed that it is to be sought for in the earth and not the atmosphere. 

 Balfour Stewart endeavors to prove that the auroras are secondary 

 currents due to small, suddeu changes in the terrestrial magnetism. 

 Heury found evidences of atmospheric electricity when snow-fall occurs 

 at one station while the heavens are clear at the other stations. 



(17) As to the relation between the aurora and the weather, numerous 

 attempts have been made to deduce some connection between the aurora 

 and the general atmosphere. After giving the belief current among the 

 natives of Arctic regions as recorded by explorers during the past hun- 

 dred years, Fritz gives in detail some of the results of more careful study. 

 Thus, Dalton finds the aurora to be a i)recursor of clear, fine weather 

 iu England. He also found a slight evidence in favor of the conclusion 

 that the barometer would rise on the day after the aurora. Hansteen, 

 from his long series of observations at Christiania, found that a lower 

 temperature almost invariably follows an aurora, at least throughout 

 Sweden, while throughout the north of Siberia the natives maintain 

 that storms of wind and rain accompany tlie aurora. Collecting these 

 and other generalizations together in tabular form one sees at a glance 

 that the apparent connection between the aurora and weather is a local 

 accident peculiar to the individual place under examination, and that 

 we are justified iu denying the existence of any influence of the aurora 

 upon the weather, but upon the other hand the influence of the weather 

 upon the aurora is not thereby denied. The relation between the 

 aurora and the temperature of the air cannot be satisfactorily proven 

 with the observations at present available. The aurora does not occur 

 at the time of the lowest temperature of the day, nor does it directly 

 follow the annual temperature changes that, according to Koppen, closely 

 correspond to the sun spots. 



(18) The relation between aurora and barometric pressure seems to ad- 

 mit of more exact description. Thus, an examination of twenty-three sta- 

 tions gives the diurnal maximum of auroras at 10 p. m., and the barometer 

 maximum at 10 hours and 20 minutes p. m. Again, the annual periods 

 of the barometer and aurora are so related that the minimum i)ressure 

 corresponds to the maximum of auroras. As regards the secular period- 

 icity of the barometer and the aurora, Hornstein, from an examination 



