parent when subjected to the action of Tyndall's electric beam. But if vapours resembling those which become transmuted into cloud under the same action existed in any part of the solar domain, there can be no doubt that his rays would render them visible precisely as the beam of the electric lamp renders visible the solid luminous spear' of Tyndall's experiment. Here, then, the fact is suggested as at least possible that comets may resemble the clouds which make their appearance when the electric light transmutes certain transparent vapours into visible nebulous matter.

And one peculiarity of comets accords well with this view. Tyndall found that when he had reduced the amount of transparent vapour in the tube to a quantity bearing an indefinitely minute proportion to the mass of the air in the same tube, the cloud still made its appearance under the action of the electric light, but was so exceedingly delicate that the faintest light seen. through it remained altogether undimmed. Now it is well known that comets present a feature precisely corresponding to this peculiarity of Tyndall's clouds. They have been known to pass over nebulæ of excessive faintness, not only without obliterating them, but without appreciably diminishing their light. This is the first of the interesting series of analogies on which Professor Tyndall's theory of comets has been founded.

According to this view, then, we are to look upon a comet as composed of a vapour which the sun's light is able to decompose: in fact, as an actinic cloud formed by the sun's decomposing power. The tail of the

comet is not matter projected from the head, either by some power inherent in the comet, or by the repulsive influence of the sun, but is matter precipitated upon the solar beams which traverse the cometary atmosphere. It must be understood, according to this theory, that the comet's atmosphere extends not only to the tail, but to an equal distance on every side of the comet's head. The sun's rays after passing through the comet are assumed to have a power which they do not ordinarily possess, the power, namely, of drawing down upon themselves from the cometary atmosphere the matter which renders them visible. Let us see how Professor Tyndall accounts for this new power.

1

A

The condensation to which the formation of the visible cloud is due he finds to depend entirely on the action of the actinic rays, and these rays are absorbed in passing through the vapour. Light-rays and heatrays have no power to produce the effects described. Nay, the heat-rays have the power of dissipating the visible cloud when the actinic rays are weakened. sort of contest may in general be supposed to be going on between the heat-rays and the actinic rays; and where one or other preponderates, there visible cloud is absent or present. Now Professor Tyndall assumes that the head and nucleus of a comet have the power of intercepting all or nearly all of the heat-rays. Hence in the part of space which is screened by the

1 In many cases this would imply that some of the longer-tailed comets have had atmospheres surrounding and including the sun and all the planets within the orbit of Mars.

head and nucleus, the actinic rays are relatively more powerful, and are thus enabled to bring down from the interplanetary spaces the matter which renders the tail visible. Elsewhere the heat-rays prevent the formation of any such visible cloudy matter.

It will be observed that this theory accounts for many facts which had seemed very perplexing. When we remember that many comets have approached the neighbourhood of the sun with a tail streaming millions of miles (in one case two hundred millions of miles) behind them, and after passing perihelion (in some instances only a few hours later), have been seen with a precisely similar tail carried in front of them, so that as Sir John Herschel remarked the apparent motion of the tail resembles that of a stick whirled around by the handle, we cannot but look with satisfaction on a theory which promises to remove so serious a difficulty. For undoubtedly the formation of a tail in one direction, and the destruction of all vestiges of former tails which had projected in other directions, would be processes which might take place with all the rapidity with which light flashes through space, if Professor Tyndall's theory be true.

Unfortunately the theory is surrounded with many and grave difficulties.

In the first place there are cometic phenomena of which it fails to give account. The formation of the luminous envelopes which the nucleus throws off as the comet approaches the sun, is a process which by no means takes place with the rapidity which Professor

Р

Tyndall's theory seems to require. I would not lay much stress on this point, however. The envelopes are frequently separated from the head of the comet by dark spaces. Now the cloudy matter existing under the conditions described by Professor Tyndall might as the comet approached the sun be in part converted by the increased heat into invisible vapour. But no sufficient reason suggests itself why this vapour after rising towards the sun should be reconverted into visible cloud. Still more perplexing (remembering always Professor Tyndall's assumption as to the nature of the vapour) seems the repetition of this process, often seen to result in the formation of several distinct envelopes.

Nor must we conceal from ourselves the fact that the appearance presented during the development of the tail is as though the matter of the envelope were being driven away by some powerful repulsive influence proceeding from the sun. It is impossible to look upon some of the drawings which experienced observers have made of comets, without feeling that processes of considerable violence are at work in the formation of the tail. I am aware that appearances of the sort are very apt to be deceptive, and therefore lay the less stress upon the evidence they afford. Still these appearances require to be considered in forming a theory of comets. There is nothing in Professor Tyndall's theory to afford any satisfactory explanation (so far as I can see) of the strange variety of forms observed in the heads, envelopes, and tails of comets.

According to this, the old tails' are dissipated by the heat-waves, so soon as these pass clear of the head towards the space occupied by the part of the tail which is to be dissipated; and Professor Tyndall accounts for the apparent bending towards the end of the tail as arising from the finite though small period occupied by the heat-waves in travelling down to the tip of the tail. It is not, therefore, the ordinary progress of the heat-waves which is in question. For heat-waves travel as fast as light-waves, and would therefore traverse the length of a comet's tail of unusually large dimensions in less than ten minutes (in which time light, as we know, would travel more than one hundred millions of miles). Hence the utmost curvature we could allow the tail from this cause is such that the direction of the tip of the tail, instead of pointing towards the actual position of the head, would point to the position the head had occupied ten minutes before. Such a deviation would be altogether inappreciable (save in one or two exceptional instances, in which, however, the contrary would only hold for a very brief interval of time); yet we know that comets' tails are often curved in a very perceptible manner, and this during the whole time of the comet's visibility. But this difficulty is removed, if, as Professor Tyndall believes, the rate at which the tail is rendered apparent (or formed, we may say) may be comparatively slow or practically instantaneous; the like holding of the rate at which the old tails are destroyed. A little consideration will show that in this