attention to this consideration. Of course a great deal must be assumed in an inquiry of the sort. Since, however, we are discussing the question whether there can be any resemblance between Jupiter and our earth, we may safely (so far as our inquiry is concerned) proceed on the assumption that the atmosphere of Jupiter does not greatly differ in constitution from that of our earth. We may further assume that at the upper part of the cloud-layers we see, the atmospheric pressure is not inferior to that of our atmosphere at a height of seven miles above the sea-level, or one-fourth of the pressure at our sea-level. Combining these assumptions with the conclusion just mentioned, that the cloud-layers are at least 100 miles in depth, we are led to the following singular result as to the pressure of the Jovian atmosphere at the bottom of the cloud-layer: -The atmosphere of any planet doubles in pressure with descent through equal distances, these distances depending on the power of gravity at the planet's surface. In the case of our earth, the pressure is doubled with descent through about 3 miles; but gravity on Jupiter is more than 2 times as great as gravity on our earth, and descent through 1 mile would double the pressure in the case of a Jovian atmosphere. Now 100 miles contain this distance (1 mile) more than seventy-one times; and we must therefore double the pressure at the upper part of the cloud-layer seventy-one successive times to obtain the pressure at the lower part. Two doublings raise the pressure to that at our sea-level; and the remaining sixty

at our sea-level so many times that the number representing the proportion contains twenty-one figures.* I say would result in such a pressure, because in reality there are limits beyond which atmospheric pressure cannot be increased without changing the compressed air into the liquid form. What those limits are we do not know, for no pressure yet applied has changed common air, or either of its chief constituent gases, into the liquid form, or even produced any trace of a tendency to assume that form. But it is easily shown that there must be a limit to the increase of pressure which air will sustain without liquefying. For the density of any gas changes proportionately to the increase of pressure, until the gas is approaching the state when it is about to turn liquid. Now air at the sea-level has a density equal to less than the 900th part of the density of water; so that if the pressure at the sea-level were increased 900 times, either the density would not increase proportionally, which would show that the gas was approaching the density of liquefaction, or else the gas would be denser than water, which must be regarded as utterly impossible.

* The problem is like the well-known one relating to the price of a horse, where one farthing was to be paid for the first nail of 24 in the shoes, a halfpenny for the next, a penny for the third, two pence for the fourth, and so on. It may be interesting to some of my readers to learn, that if we want to know roughly the proportion in which the first number is increased by any given number of doublings, we have only to multiply the number of doublings by ths, and add 1 to the integral part of the result, to give the number of digits in the number representing the required proportions. Thus multiplying 24 by ths gives 7 (neglecting fractions); and therefore the number of farthings in the horse problem is represented by an array of 8 digits.

Or if any one is disposed, for the sake of argument, to assert that a gas (at ordinary temperatures) may be as dense as water, then we need proceed but a few steps farther, increasing the pressure about 18,000 times instead of 900 times, to have the density of platinum instead of that of water, and no one is likely to maintain that our air could exist in the gaseous form with a density equalling that of the densest of the elements. We are still an enormous way behind the number of twenty-one figures mentioned above; and in fact, if we supposed the pressure and density to increase continually to the extent implied by the number of twenty-one figures, we should have a density exceeding that of platinum more than ten thousand millions of millions of times!

Of course this supposition is utterly monstrous, and I have merely indicated it to show how difficulties crowd around us in any attempt to show that a resemblance exists between the condition of Jupiter and that of our earth. The assumptions I made were sufficiently moderate, be it noticed, since I simply regarded (i.) the air of Jupiter as composed like our own; (ii.) the pressure at the upper part of his cloud-layer as not less than the pressure far above the highest of our terrestrial cumulus clouds (with which alone the clouds of Jupiter are comparable); and (iii.) the depth of his cloud-layer as about 100 miles. The first two assumptions cannot fairly be departed from to any considerable extent, without adopting the conclusion that the atmosphere of Jupiter is quite

to maintain. The third is, of course, open to attack, though I apprehend that no one who has observed Jupiter with a good telescope will question its justice. But it is not at all essential to the argument that the assumed depth of the Jovian atmosphere should be even nearly so great. We do not need a third of our array of twentyone figures, or even a seventh part, since no one who has studied the experimental researches made into the condition of gases and vapours can for a moment suppose that an atmosphere like ours could remain gaseous, except at an enormously high temperature, at a pressure of two or three hundred atmospheres. Such a pressure would be obtained, retaining our first two assumptions, at a depth of about fourteen miles below the upper part of the cloud-layer. This is about the 6,000th part of the diameter of Jupiter; and if any student of astronomy can believe that that wonderfully complex and changeful cloud envelope which surrounds Jupiter, has a thickness of less than the 6,000th part of the planet's diameter, I would recommend as a corrective the careful study of the planet for an hour or two with a powerful telescope, combined with the consideration that the thickness of a spider's line across the telescopic field of view would suffice to hide a breadth of twenty miles on Jupiter's disc.

But we are not by any means limited to the reasoning here indicated, convincing as that reasoning should be to all who have studied the aspect of Jupiter with adequate telescopic power. We have in Jupiter's mean density an argument of irresistible force against the only view which

enables us even hypothetically to escape from the conclusions just indicated. Let it be granted, for the sake of argument, that Jupiter's cloud-layer is less than fourteen miles in depth, so that we are freed for the moment from the inference that at the lower part of the atmosphere there is either an intense heat or else a density and pressure incompatible with the gaseous condition. We cannot, in this case, strike off more than twenty-eight miles from the planet's apparent diameter to obtain the real diameter of his solid globe-solid, at least, if we are to maintain the theory of his resemblance to our earth. This leaves his real diameter appreciably the same as his apparent diameter, and as a result we have the mean density of his solid globe equal to a fourth of the earth's mean density precisely as when we leave his atmosphere out of the question. Now I apprehend that the time has long since passed when we can seriously proceed at this stage to say, as it was the fashion to say in text-books of astronomy, "therefore the substance of which Jupiter is composed must be of less specific gravity than oak and other heavy woods." We know that Brewster gravely reasoned that the solid materials of Jupiter might be of the nature of pumice stone, so that with oceans resembling ours a certain latitude was allowed for increase of density in Jupiter's interior. But in the presence of the teachings of spectroscopic analysis, few would now care to maintain, as probable, so preposterous a theory as this. Everything that has hitherto been learned respect