CLASSIFICATION OF COMETS. 
359 
circumstances whatever, a particle crossing the track of Uranus 
without encountering the planet could be deprived of 4-J miles 
per second of its velocity. For though Uranus can deprive 
a body directly receding from him (and starting from his 
surface) of a velocity of about 13 miles per second, yet the con- 
siderations above adduced show that only a fraction of this 
velocity could be abstracted from a body moving past Uranus ; 
and it is certain that if so large a reduction as 4J miles per 
second could be effected at all, it would only be by a singularly 
close approach of the particle to the surface of Uranus. 
But setting apart the improbability that a body arising from 
interstellar space could be in this way compelled to travel in 
the orbit of the November meteors, the possibility of such a 
capture would not prove the possibility of the capture of a flight 
of bodies large enough to form that meteor system and its 
accompanying comet. If the whole material of the system and 
its comet had arrived in a compact body, the material attrac- 
tions of the parts of that body would be sufficient to keep them 
together ; whereas, in point of fact, the November meteor system 
and its comet occupy at present a large range of space, even if 
the meteors be not scattered all round the orbit (however 
thinly along portions thereof). If, on the other hand, the 
material of the body were not in a compact form, the body 
would be necessarily large, and a portion of it only would be 
captured by Uranus. Nay, it is not even necessary that this 
should be conceded. For though we admitted that the whole 
of a large and tenuous body not kept together by the mutual 
attraction of its parts or by cohesion, might be captured, it is 
manifest that different parts would be captured in different 
w 7 ays, and would thenceforth travel on widely different orbits. 
That a system of bodies already drawn out into an extended 
column, and in respect of length already resembling the meteor 
systems we are acquainted with, could be captured, as Schiapa- 
relli’s theory requires, and all sent along one and the same 
closed orbit, is altogether impossible. 
It is to be noticed also that we gain nothing, as respects the 
interpretation of comets, by adopting Schiaparelli’s hypothesis. 
To assume that cometic matter has been wandering about 
through interstellar space, until the sun’s attractive influence 
drew such matter towards the solar system, is to explain a 
difficulty away by advancing another still greater ; moreover, 
we have not a particle of evidence in support of the supposition. 
To suppose, on the other hand, that comets have crossed the 
interstellar spaces, coming to us from the domain of another 
sun, is to remove the difficulty only one step. We know that 
comets pass away from the domain of our sun to visit some 
other sun after an interstellar journey of tremendous duration ; 
