existencc. But even in the case of our solar system, though the evidence in some cases against the possibility of life is exceedingly strong, we do not meet with a single instance in which evidence of the contrary kind is forcible, still less decisive. So that in the solar system the evidence is almost as clear in favour of the conclusion above indicated as where we reason about worlds of whose actual condition we know nothing. As respects such worlds, that is, as respects the members of those systems of worlds which circle, as we believe (from analogy), around other suns than ours, the probability that any particular world is inhabited at this present time is exceedingly small.

But let us next consider what is the probability that there is life on some member or other of a scheme of worlds circling around any given sun. Here, again, the argument is from analogy, being derived from what we have learned or consider probable in the case of our own system. And I think we may adopt as probable some such view as I shall now present. Each planet, according to its dimensions, has a certain length of planetary life, the youth and age of which include the following eras:-a sunlike state; a state like that of Jupiter or Saturn, when much heat but little light is evolved; a condition like that of our earth; and lastly, the stage through which our moon is passing, which may be regarded as planetary decrepitude. In each case of world-existences the various stages may be longer or shorter, as the whole existence is longer or shorter, so that speaking generally the period of habitability bears

the same proportion in each world to the whole period of its existence; or perhaps there is no such uniform proportion, while, nevertheless, there exists in all cases that enormous excess of the period when no life is possible over the period of habitability. In either case, it is manifest that regarding the system as a whole, now one, now another planet (or more generally, now one, now another member of the system) would be the abode of life, the smaller and shorter-lived having their turn first, then larger and larger members, until life has existed on the mightiest of the planets, and even at length upon the central sun himself. We need not concern ourselves specially with the peculiarities affecting the succession of life in the case of subordinate systems, or of the members of the asteroidal family, or in other cases where we have little real knowledge to guide us: the general conclusion remains the same, that life would appear successively in planet after planet, step by step from the smaller to the larger, until the approach of the last scene of all, when life would have passed from all the planets, and our sun would alone remain to be in due time inhabited, and then in turn to pass (by time-intervals to us practically infinite) to decrepitude and death.

During all this progression, the intervals without life would in all probability be far longer than those when one or other planet was inhabited. In fact, the enormous excess of the lifeless periods for our earth over the period of habitability, renders the conclusion all but certain that

last very much longer than the periods of life (in this or that planet) with which they would alternate.

If we apply this conclusion to the case of any given star or sun with its scheme of dependent worlds, we see that even for a solar system so selected at random the probability of the existence of life is small. It is, of course, greater than for a single world taken at random ;—just as if I had ten friends who were to be at home each for six minutes between noon and ten, the chance would be greater that some one of the number would be at home at a given moment of that interval than would be the chance that a given one of the number would be then at home; while yet even taking all the ten it would still be more likely than not that at that moment not one would be at home.

Thus when we look at any star, we may without improbability infer that at the moment that star is not supporting life in any one of those worlds which probably circle round it.

Have we then been led to the Whewellite theory that our earth is the sole abode of life? Far from it. For not only have we adopted a method of reasoning which teaches us to regard every planet in existence, every moon, every sun, every orb in fact in space, as having its period as the abode of life, but the very argument from probability which leads us to regard any given sun as not the centre of a scheme in which at this moment there is life, forces upon us the conclusion that among the millions on millions, nay, the millions of millions of suns which

people space, millions have orbs circling round them which are at this present time the abode of living creatures. If the chance is one in a thousand in the case of each particular star, then in the whole number (practically infinite) of stars, one in a thousand has life in the system which it rules over: and what is this but saying that millions of stars are life-supporting orbs? There is then an infinity of life around us, although we recognise infinity of time as well as infinity of space as an attribute of the existence of life in the universe. And remembering that as life in each individual is finite, in each planet finite, in each solar system finite, and in each system of stars finite, so (to speak of no higher orders) the infinity of life itself demonstrates the infinity of barrenness, the infinity of habitable worlds implies the infinity of worlds not as yet habitable, or which have long since passed their period of inhabitability. Yet is there no waste, whether of time, of space, of matter, or of force; for waste implies a tending towards a limit, and therefore of these infinities, which are without limits, there can be no waste.

A MISSING COMET.

MANY persons were alarmed in August 1872 lest it should be true (as reported) that Plantamour, the Swiss astronomer, had predicted the earth's destruction by a comet on the twelfth of that month. When once a prediction of this sort has been announced, it is almost impossible to remove the impression produced by it. The reputed author of the prediction may deny flatly that he had ever announced even the approach of a comet; every astronomer of repute may add his testimony to the effect that no comet is due at the time indicated for the earth's destruction; the way in which the mistake arose may be explained, and every effort made to spread the explanation as widely as possible: yet the impression will nevertheless remain that there must have been some ground for the prediction, or if it be insisted that no prediction was made then there must have been some ground for the story of the prediction. Confidence is not completely restored until the day and hour announced for the earth's destruction have passed without mishap.*

Being at Sheffield in October 1872, I was told an excellent story about the comet. The story has the advantage over most others of the kind, of being strictly true :-In a certain house, in Sheffield, Monday, August 12, had been appointed a great washing-day. On the morning