HISTORY OF PHYSICAL ASTRONOMY.

309

radius vector should always be in the same direction when the cometappears in the same region of the heavens. Newton, however, hasremarked that, although the comet of 1577 occupied the same apparentposition on tire 28th of December which the comet of 1680 occupied onthe 29th of the same month of the year, and although the earth, in bothcases, was consequently in the same part of her orbit, the tail of thecomet of 1577 deviated 21° towards the south, whereas on the other handthe tail of the comet of 1680 deviated 4|° towards the north*. It follows,therefore, that the supposition of the phenomenon being due to therefraction which the light of the comet suffers in passing through thecelestial regions is inconsistent with observed facts, and is consequentlyuntenable. Another opinion respecting the tails of comets is that whichsupposes them to be appearances similar to the aurora borealis. A theory,founded upon this view of the subject,.was proposed by Mairan, but itsinconsistencies were too glaring to secure even its partial adoption.

According to the various hypotheses hitherto considered, the tail of acomet is merely an optical illusion, having no real foundation in nature.Another class of hypotheses is that according to which they are imaginedto be composed of a material substance similar to the nebulosity of thecomet. Some persons, adopting the maxims of the Aristotelian philosophy,respecting light and heavy bodies, supposed that the matter composingthe atmosphere of comets w r as au essentially light substance; that invirtue of its inherent levity it had a constant tendency to recede from thesun; and that hence originated the appendage of the tail. It is hardlynecessary to state that this hypothesis is at variance with the fundamentalproperty of the inertia of matter and the recognised principle of universalgravitation.

The first approach to anything resembling a rational explanation of thetails of comets is unquestionably due to Kepler. According to thatillustrious astronomer a comet is composed wholly of a nebulous substance,the constituent parts of which are gradually broken and dispersed by theincessant action of 'the solar rays upon them. The lighter particles yieldto the impulse of the rays, and, proceeding to an immense distance fromthe head of the comet, occasion the appendage of the tail. The denserparticles remain behind and form the nebulosity surrounding the head-)-.This hypothesis furnished a satisfactory account of the general directionof the tails of comets. It also afforded an explanation of their curvatureand their concavity wdth respect to the region which the comet wasleaving; for it is manifest that as the comet continued to revolve in itsorbit, the nebulous particles impelled by the solar rays would lag some-what behind the prolongation of the radius vector, and that the effect soproduced would be greatest for the particles which had ascended theearliest, or, in other words, for the particles most remote from the head.Another advantage of this hypothesis was that of referring the phenomenonto a true physical cause; for it can hardly be doubted, whatever be themode of the propagation of light, that the solar rays, if interrupted intheir progress by a material substance, must communicate to it an impactof some degree of intensity. On the other hand it must he admitted thatthe cause assigned, although founded in nature, cannot, with any degreeof probability, according to Kepler’s view of it, be considered sufficient toproduce the observed appendage of the comet. A material improvement

Princip., book iii., prop. 41.

t De Cometis, p. 100.