THE ILLUMINATION OF THE ECLIPSED MOON. 



3G1 



tho atinoaplioric rinp;, wo find tliat in the middle of tlio 

 former eclipse tlio proportion of land was very much less, 

 and of water very much more, than in the middle of the 

 latter eclipse. If, therefore, any effect at all is to be 

 ascribed to this cause, tho eclipse of March, 1848, should, 

 on the more reasonable view of tho effect of vapour, have 

 been a comparatively dark eclipse, and the eclipse of 

 October, 1884, shoiild have been a comparatively bright 

 ecliiiHO. We know the facts to have been the exact reverse 

 of this. A"-ain, if we compare the eclipse of October, 1884, 

 with the eclipse of .Tanuary last, we find that the proportion 

 of land to water at the time of mid-eclipse was almost 

 identical in. the two eclipses. Yet in the one case the moon 

 was so dim as to be practically invisible ; in tho other it 

 was at least of average brightness. These results were 

 obtained by careful experiments with a terrestrial globe. 



It has been supposed by some that the eclipse of October, 

 1R84, may luivo owed its remarkable darkness to tho pre- 

 sence in tho eartli's atmosphere of a vast quantity of 

 volcanic dust from the Krakatoa eruption of tho previous 

 year, and that the other dark eclipses on record may admit 

 of a similar explanation. I will only say of this theory that 

 the cause seems to me inadequate to tho effect, and that if 

 the atmosphere generally had been charged with volcanic 

 dust to the extent that the theory requires, more decided 

 evidence of the presence of this material might have boon 

 expected than any we possess. In particular, we might 

 have expected to find, after a sufficient lapse of time, 

 copious deposits of dust over large tracts of the earth's 

 surface. 



The theory of the atmosphere being charged with cosmic 

 or meteoric dust, or dust of extra-terrestrial origin, is, I 

 think, equally open to the last-named objection. 



