CLASSIFICATION OF COMETS. 
351 
have no more characteristic difference to deal with than that of 
dimensions. Supposing, for instance, that we could separate 
comets into those with or without a nucleus, or those with or 
without a tail ; such a classification, if it was found to correspond 
with a real difference of nature, would be much more satisfac- 
tory than the arrangement of comets into various orders differ- 
ing only in size. One of the most interesting questions, then, 
in the cometic astronomy of a few years ago was this — Are the 
peculiarities just referred to — the absence or presence of a 
nucleus, or of a tail — really characteristic, or do they correspond 
to mere differences of development ? I say that this question 
belonged to cometic astronomy of a few years ago, though even 
then there were reasons for regarding the various forms of 
structure observed in comets as depending only on development. 
Of course comets which, during the whole time of their visibility, 
showed neither tail nor well-defined nucleus, could afford no 
means of answering the question. But a comet like Donati’s — 
the glorious plumed comet of 1858 — which appeared as a mere 
globular haze of light, and gradually during its approach to the 
sun assumed one form after another of cometic adornment — the 
nucleus, the fan-shaped expansion, the long curved tail, stria- 
tions within the tail and envelopes outside the fan, while finally 
even subsidiary tails made their appearance — teaches us unmis- 
takably that these features depend merely on development. We 
might as reasonably place the chicken in another class than the 
full-grown fowl because it has neither comb nor coloured tail- 
feathers, as set a small comet in another order than that to 
which Donati’s belongs because the small one shows neither 
tail nor coma. The gradual loss of these appendages by 
Donati’s comet, during its retreat into outer space, of course 
strengthens this view. But perhaps the most remarkable proof 
ever afforded of the variety of appearance which the same comet 
may present, was that given by Halley’s comet at its return in 
1835-36; for on that occasion, after showing a fine coma and 
tail during its approach towards the sun, it was seen in the 
southern hemisphere by Herschel and Mac! ear, not only without 
tail, but even without coma, appearing in fact precisely like a 
star of the second magnitude. After this — that is to say, during 
its retreat — it gradually resumed its coma, and even seemed to 
be throwing out a new tail, but no complete tail was formed 
while the comet remained visible. 
Indeed the difference between the appearance presented by 
the same comet before and after its nearest approach to the 
sun is not only remarkable in itself, but subject to remarkable 
variations. 66 What is very remarkable,” says Sir John Herschel 
on the first point, “ the shape and size are usually totally dif- 
ferent after the comet’s reappearance (on the other side of the 
