354 
POPULAR SCIENCE REVIEW. 
our sun on such a path, the very least of the multitudinous 
disturbing attractions to which the comet would be exposed 
would suffice to change the path either to the elliptic or the 
hyperbolic form. 
And here we may pause to inquire how far the second of the 
three classes into which comets have thus been divided can be 
regarded as a class apart. Does the mere fact that a comet has 
a re-entering path — so that, unless perturbations affect it, the 
comet will travel in continual dependence on our sun — afford a 
sufficient reason for distinguishing the comet from others which 
travel on a hyperbolic path ? It appears to me that this 
question admits of being answered in two ways. When we 
remember that a comet approaching our system on a slightly 
hyperbolic path might have that path changed into an elliptic 
figure by the perturbations to which the comet would be subjected 
during its visit, we may reasonably decide that the mere fact of 
a comet pursuing an elliptic path ought not to be considered a 
valid reason for distinguishing it from one of the hyperbolic 
comets. But when we consider, on the other hand, that there 
are comets like those of Jupiter’s family, which are quite dis- 
tinctly separated by the nature of their paths from the hyperbolic 
comets, we may not unreasonably infer that some at least of 
those which travel on elliptic paths of great eccentrity are in 
reality to be classified apart from the hyperbolic comets, as 
having had a different origin and a different history. We might, 
indeed, reverse the argument just adduced, and reason that the 
hyperbolic comets ought not to be classified apart from the 
comets of long period, because perturbations excited within the 
solar system might change an elongated elliptic orbit into a 
hyperbolic one. The point at issue is thus seen to resolve itself 
into the question whether we can assert that there are comets 
which from the earliest times (the youth of the solar system) 
have belonged to it (i.) with short periods and (ii.) with long 
periods, while (iii.) other comets have visited it from other sys- 
tems. We find in fact that the attempt to classify leads in 
this case, as it has led in so many others (as perhaps it inevitably 
must lead, if properly conducted), to the question of origin. 
And here perhaps the question will arise, may we not cut the 
Gordian knot by denying that even the comets of short period can 
be separated from the hyperbolic comets which visit our system 
from interstellar space ? I am aware that the theory of comets 
and meteors which Schiaparelli has advanced, and which many 
in this country have viewed with considerable favour, points to 
this conclusion. For according to that theory meteor-systems 
are groups of discrete bodies which have been drawn towards 
our solar system, gradually lengthening out as the process of in- 
draught continued, and have then been compelled by the 
