48 • Mr. G, J. Stoney [Feb. 14, 



would now and then pass in various directions, and with various velo- 

 cities through its substance. For the most part they would go entirely 

 through and pass out again ; but in every such case the meteor would 

 leave the comet with less velocity than it had when approaching it. 

 And in some cases this reduced velocity would be such that the future 

 path of the meteor would be an ellipse round the comet. Whenever 

 this was once brought to pass, the meteor would inevitably return 

 again and again to the comet, each time passing through some part of 

 its substance, and at every passage losing speed. After each loss of 

 speed the ellipse it would next proceed to describe would be smaller 

 than the one before, until at last the meteor would sink entirely into 

 the gas and be engulfed by it. In this way meteor after meteor would 

 settle down through the comet, and, in the end, just such a cluster 

 would be formed as came across the planet Uranus in the year 126, 

 or, if such a cluster existed originally within the mass of gas, it would 

 in this way be augmented. As the comet swept past the planet, its 

 outlying parts would seem to have grazed his surface, and in this way 

 the gas was probably somewhat more retarded than the meteors ; and 

 in the centuries which have since elapsed the meteors have gone so 

 much ahead of the comet that they are now treading on his heels and 

 on the point of overtaking him, while probably the gas has again 

 brought together a smaller cluster of the meteors. 



PROFESSOR GRAHAM. 



The question now arises. How the deserts of space which extend 

 from star to star come to be tenanted here and there by a patch of 

 gas or an occasional meteorite ? Light has been thrown on this in- 

 quiry by discoveries made with the spectroscoj)e in modern times and 

 by observations during eclipses. These have revealed to us the fact 

 that violent outbursts occur ujDon the sun, and doubtless on other 

 stars, so swift that the up-rush must sometimes carry matter clear 

 away into outer space. Imagine such a mass consisting in part of 

 fixed gas and in part of condensable vapours ejected from some star. 

 As it travels forward the vapours cool into meteorites, while the fixed 

 gas spreads abroad like a great net, to entangle other meteors. In 

 some cases both might travel together ; in others the gaseous portion 

 would be retarded before it passed beyond the neighbourhood of the 

 star, and the denser meteors would get ahead. But even so in the 

 lapse of ages other meteors would be caught, so that in any event a 

 cluster would at length be formed. Now, the reasonable suspicion 

 that this is the real origin of meteors has received striking confirma- 

 tion from the discovery of the late Professor Graham, that meteoric 

 iron contains so much hydrogen occluded within it as indicates that 

 the iron had cooled from a high temperature in a dense atmosphere of 

 hydrogen — precisely the conditions under which the vapour of iron 

 would cool down while escaping from a large class of stars, including 

 our sun. 



