owe their origin to the difference of absolute rotational velocity between the equatorial parts of the earth and parts in high latitudes. In the case of Jupiter the difference of this kind is not sufficient to account for the observed belts—partly because there are many, partly because they are variable, but principally because Jupiter is so much larger than the earth that 'much greater distances must be traversed ia passing from any given latitude to another where the rotational velocity is so many miles per hour more or less. Combining with these considerations the circumstance that the solar action which causes the atmospheric movements from one latitude to another in the case of our earth is reduced to one twenty-seventh part only of its terrestrial value in the case of Jupiter, we must clearly look to some other cause for the difference of absolute rotational velocity necessary to account for the belts of Jupiter.
Now, it seems to me that we are thus at once led to the conclusion that the cloud-masses forming the belts of Jupiter are affected by vertical currents, up-rushing motions carrying them from regions nearer the axle, where the absolute motion due to rotation is slower, to regions farther from the axis, where the motion due to rotation is swifter, and motions of down-rush carrying them from regions of swifter to regions of slower rotational motion. This view seems certainly encouraged by what we find when we come to study more closely the aspect of the Jovian belts. The white spots—some small, some large—which are seen to form from time to time along the chief belts present precisely the appearance which we should expect to find in masses of vapor flung from far down below the visible cloud-surface of Jupiter, breaking their way through the cloud-layers, and becoming visible as they condense into the form of visible vapor in the cooler upper regions of the planet's atmosphere. Then, again, the singular regularity with which in certain cases the great, rounded white clouds are set side by side, like rows of eggs upon a string, is much more readily explicable as due to a regular succession of up-rushes of vapor, from the same region below, than as due to the simultaneous up-rush of several masses of vapor from regions set at uniform distances along a belt of Jupiter's surface. The latter supposition is indeed artificial and improbable in the highest degree, and in several distinct respects. It is unlikely that several up-rushes should occur simultaneously, unlikely that regions whence up-rush took place should be set at equal distances from each other, unlikely that they should lie along the same latitude parallel. On the other hand, the occurrence of up-rush after up-rush from the same region of disturbance, at nearly uniform intervals of time, is not at all improbable. The rhythmical succession of explosions is a phenomenon, indeed, altogether likely to occur under certain not improbable conditions—as, for instance, when each explosion affords an excess of relief, if one may so speak, and is therefore followed by a reactionary process, in
