THE EARTH IN METEORIC SHADOW. 159 



secondly, that every existing comet is followed by a 

 train of meteors (which we may well believe to be in 

 some degree proportional in wealth of distribution and, 

 as respects the size of its components, to the comet 

 with which it is associated). Now it is further found 

 that the number of cometic paths increases rapidly as 

 we approach the sun; and not only is the rate of 

 increase rapid, but this rate is itself increasing as the 

 sun is approached, so that whatever the wealth of 

 cometic, and therefore of meteoric distribution, at the 

 earth's distance from the sun, the wealth of such dis- 

 tribution close by the sun is enormously greater. 



But we have other evidence on this point which, 

 though perhaps by itself it might not be very weighty, 

 amounts almost to demonstration when regarded as 

 interpreting, and also as interpreted by, the evidence 

 we have just been dealing with. 



The solar corona might in former days have been 

 regarded as in no way connected with the subject we 

 are considering ; for in former days, despite the really 



recognised by annual star-showers was due to meteoric bodies fol- 

 lowing in the track of a comet already dissipated, viz., Eiela's (or 

 Gambart's, as it is more properly called). This comet, which was 

 divided into two in 1846, was looked for in vain in 1866, and when 

 next due, in 1872; but because then it should have passed the earth's 

 track (which it chances nearly to cross) shortly before the earth arrived 

 near the passing-place, some students of astronomy (I was one of them, 

 and, I believe, the first) suggested that, in all probability, when the 

 earth arrived at that part of her orbit about November 27, there would 

 be a display of meteors, radiating (because following the comet) from 

 a part of the star-sphere near the feet of Andromeda ; and such a dis- 

 play was seen (one of the richest star-showers of the century, numeri- 

 cally) on the night of November 27, and with the predicted ' radiant 

 point.' 



