304 METEOES AND AEROLITES. 



and justifies this by the record of results which show that 

 other and new conditions must be added to the theories of 

 meteoric phenomena now received. Of the more remarkable 

 of these results we shall give a short summary ; such as may 

 enable our readers to judge of their nature and bearing on 

 the argument. 



Observations on shooting-stars and other meteors were 

 begun by M. Coulvier-Gravier at Eheims as early as 1811, 

 under electrical and other theories of their origin, which he 

 afterwards abandoned. It was not, however, until 1841 that, 

 at the suggestion of Arago, he began carefully to register 

 their number, times of appearance, and direction in the 

 heavens. In 1845 M. Saigey associated himself to his 

 labours, and aided greatly in generalising and giving method 

 to the results. In a period of 42 months, between 1841 

 and. 1845, there were 5,302 shooting-stars recorded; seen 

 during 1,054 hours of observation. The number would 

 doubtless have been much greater but for the interference of 

 the moon, which, when full, effaces nearly three-fifths of the 

 stars otherwise visible. An estimate made, with allowance 

 for this cause, brings out the mean horary number of 6 ; the 

 actual mean number seen per hour being 5 -6. The passing 

 obscuration by clouds makes another void in the calculation, 

 the amount of which it is difficult to estimate. 



But this general horary mean loses its interest in another 

 more curious and unlooked-for result of these observations, 

 viz. the variations found to exist at different hours. With 

 rare exceptions, the number of visible meteors increases as 

 the night advances ; and this at all times of the year, and 

 with regularity enough to furnish the basis of tables for each 

 successive hour of the night. A few instances we give from 

 different hours between evening and morning. In the 

 evening from 6 to 7 o'clock the mean number of stars falling 

 is 3-3 from 9 to 10 o'clock 4 from 11 to 12 o'clock 5 



