1889.] . of the various Species of Heavenly Bodies. 211 



Further, it is important to associate the spectra of the envelopes 

 and nucleus with the multiplicity of tails. 



Let us suppose a comet's tail thus chemically constituted; the 

 molecules will be moving rapidly under the influence of the solar 

 repulsion away from the meteorites which produce them, through a 

 meteoritic plenum. Hence we should expect auroral phenomena. 

 These have been recorded in comets' tails since the time of Kepler. 

 In the tail we have gases moving through meteoritic dust, in the 

 aurora, as I shall show in the next part of this memoir, we have in all 

 probability meteoritic dust moving through gases. 

 What then becomes of the tails ? 



Being thus formed at the expense of the materials composing the 

 head, the materials removed from the head can never be returned to 

 it because of its insufficient gravitational power over them, and more- 

 over they can no longer traverse the same orbits as the meteorites 

 from which they sprung, because they have already been turned out 

 of that course by the forces attending the development of the tail. 

 The gaseous bodies thus become distributed throughout the space 

 occupied by our system, and give no further trace of their existence 

 until, after subsequent occlusion which causes their disappearance, 

 they are again made evident by future collisions. The existence of 

 " unperihelioned matter " then indicates that the regions of space 

 nearer the sun are not so full of these free gaseous products as those 

 further away. 



Comets must thus degenerate, so far at all events as their easily 

 volatilised constituents are concerned, with each perihelion passage, 

 but as the majority of them only approach the sun at long intervals 

 of time they do not suffer much in this way. Some of the short- 

 period comets get less and less brilliant at each successive perihelion 

 passage, and others are then observed entirely without tails, all the 

 available tail-forming material having been used up and dispersed 

 into the regions of space farther away from the sun, while at aphelion 

 a fresh supply has been lacking. 



It has been conjectured by Weiss and Schiaparelli that the con ,. 

 densed metallic materials of the tails, which are projected with the 

 tails in the cases of the comets whose perihelia lie within the earth's 

 orbit, may give rise to the appearance of meteors. 



This may also happen in the case of condensable materials shot in 

 the first instance towards the sun, so that we may imagine the original 

 train of meteorites to gradually widen out in the plane of the orbit 

 inside and outside of the orbit of the main swarm.* 



It has been suggested that the luminosity of comets is possibly 

 partly electrical, and in support of this view Hasselberg showed that 

 the changes in Wells's Comet were closely related to changes which 

 * Herschel, ' Monthly Notices/ vol. 35, p. 253. 



