MoNCK — On Star- Distribution. 503 



ninth sphere, as computed from the number of stars, is 21 '54, while 

 that computed from the total light is only 18-82, This would 

 seem to indicate a loss of at least 12|^ per cent, of the total light 

 of all the stars up to and including those of the ninth magnitude, 

 which of course implies a much greater loss in the case of an 

 average ninth magnitude star ; and if I had made the computation 

 between the second and ninth spheres instead of the first and 

 ninth, the loss would appear to be over 35 per cent. 



As our knowledge of the ether increases, I think the chances of 

 its proving a light-absorbing medium are becoming greater, on 

 purely physical grounds, and apart from astronomical data. We 

 know of no other medium which transmits vibrations without ab- 

 sorbing some of them. The ether is an active agent in the produc- 

 tion of electric and magnetic phenomena, and thereby, no doubt, 

 assists in the production of light and heat otherwise than by trans- 

 mission. Independently too of the supposed evidence of Encke's 

 comet, there are magnetic phenomena in which the ether, when in 

 a particular condition, appears to be capable of resisting the motion 

 of ponderable matter. There seems, moreover, to be some reason 

 for thinking that all bodies in the universe (except a small number, 

 which are temporarily heated by collision or some similar cause) 

 are perpetually cooling ; and if so, what becomes of the heat unless 

 it is absorbed by the ether? However, I shall not pursue this 

 subject farther. 



The absorption or non-absorption of light will of course make 

 a considerable difference in our estimates of the distances of very 

 faint stars. From the foregoing figures it would seem that the 

 average distance of a star in the ninth sphere is about twenty-one 

 and a-half times that of a star in the first sphere. The ninth sphere, 

 however, includes not merely the stars of the ninth magnitude, but 

 those of all magnitudes higher than the ninth, so that the average 

 distance of a ninth magnitude star will be greater than this. It 

 cannot, however, be very much greater, because 71 per cent, of the 

 total number of stars included in the ninth sphere are stars of the 

 ninth magnitude. If we assume it to be twenty-six times as great 

 as that of a first magnitude star, we shall probably have made a 

 sufficient allowance. On the hypothesis of uniformity, however, 

 and not allowing for any absorption, the ratio would be (i^/lO)^ 

 to one, or very nearly thirty-nine times as great. I have already 



