all good observing telescopes. Now, of course, the satellite did not really stop. Nothing short of a miracle could have stopped the satellite, or, if the satellite could have stopped, have set it travelling on again as usual. For the satellite did not lose one mile, or change its velocity by the thousandth part of a mile per hour or even per annum. But suppose such a change had taken place at the edge of Jupiter as we have seen would certainly have taken place there if the changes affecting the spot which South saw had occurred to a region at the edge, as in fig. 26, instead of the middle, as in fig. 25. Then Smyth's observation would be perfectly explained. We require, indeed, to suppose the change occurring in a different order, the outer cloud-layer being in the first instance well-developed and very rapidly becoming dissipated, so that the outline which had at first been at its usual level, was very rapidly depressed to the inner cloud-layer. But, of course, if the rapid formation of clouds by condensation can occur on Jupiter, so also can the rapid dissipation occur, especially at that particular part where Smyth saw the satellite behave so strangely. For that part is being carried, by the planet's swift rotation, into sunlight, and the extra heat to which it is thus exposed might readily effect the dissipation of widely extended cloud strata, supposing the temperature near that critical value at which clouds form or are dissipated. Here, then, is an explanation of a phenomenon which otherwise seems utterly inexplicable. The explanation requires only that a process like one which has been observed to occur on Jupiter's disc should occur at a part of his surface forming at the moment a portion of his outline. If we had never known of such changes as South and other observers have noted in the markings of Jupiter, we should be compelled by Smyth's observation to admit their possibility. If we had never known of Smyth's observation we should be compelled by South's to admit that such a change of outline as is indicated by Smyth's observation must be possible,--must, in fact, occur whenever cloud-masses form or are dissipated over wide areas at the apparent edge of the planet. When we have both forms of evidence it seems altogether unreasonable to entertain any further doubt on this point.