Of course, it is just possible that a great dark rift, such as I have described, might appear thus to change in position without any actual transference of the bordering cloud-masses. Mr. Webb, speaking of a number of phenomena, of which those presented by the great rift of 1860 are but a few, says that "they prove an envelope vaporous and mutable like that of the earth, without, however, necessarily inferring" (? implying) "the existence of tempestuous winds: even in our own atmosphere, when near the dew-point, surprising changes sometimes occur very quietly: a cloud-bank observed by Sir John Herschel, April 19, 1827, was precipitated so rapidly that it crossed the whole sky from east to west at the rate of at least 300 miles per hour; and alterations far more sudden are conceivable where everything is on a gigantic scale." It does not seem to me altogether probable that more rapid alterations would affect cloud-banks covering millions of square miles than occasionally affect terrestrial cloud-banks covering perhaps a few tens of thousands of square miles; on the contrary, as small terrestrial clouds change relatively in a far more rapid way than large ones, and these than cloud-masses covering a county or a country, so it would seem that the changes affecting our largest cloud-layers would be relatively far more rapid than those affecting cloud-masses which could (many times over) enwrap the whole frame of this earth on which we live. But apart from that, and apart also from the important consideration that all such processes as evaporation and condensation, so far as the sun brings them about, should proceed far more sluggishly in the case of a planet like Jupiter than in that of our earth, which receives some twenty-seven times as much heat from the sun (mile for mile of surface), it is utterly incredible that precipitation should have occurred so steadily and swiftly along one edge of the great rift, and condensation—with such exactly equal steadiness and swiftness—on the opposite edge, that, while the rift as a whole shifted its position during a hundred Jovian nights and days at the rate of 150 miles per hour, its sides should nevertheless remain parallel all the time. Such processes may be spoken of as possible, in the same sense that it is possible that a coin tossed fifty times in succession should always show the same face; but we do not reckon such possibilities among scientific contingencies.
But the motion of great rounded masses in the atmosphere of Jupiter is still more decisive as to the existence not only of a very deep atmosphere, but also as to the swift motions taking place in that atmosphere.
I would, in the first place, note that the very existence of belts in the Jovian atmosphere, and especially of variable belts, implies the great depth at which the real surface of the planet must lie below the visible cloud-layers. Atmospheric belts can only be formed where there are differences of rotational velocity. In the case of our own earth we know that the trade-wind zone and the counter-trade zone
