Let _a b_, fig. 25, represent the great dark space seen by South, just below the principal belt, and let us suppose the planet turned round until this dark space, or rather this opening in the planet's outer cloud-envelope, is brought to the edge as at _a c d_, fig. 26. Then this opening would really cause a depression in the planet's outline at _d c_, the shaded part being depressed. The depression might not be observable in any ordinary telescope. For at the edge of Jupiter the features of the belt are generally lost, and the outline is at all times smoothed in appearance by that peculiarity of vision which makes all bright objects seen on a dark background appear somewhat larger than they actually are. (This is _really_ due to a fringing, as it were, of the image on the retina of the eye.) But though the depression might not be recognisable, it would exist, and, as we shall presently see, it might be detected in another way than by being actually seen. When the clouds formed which concealed the spot,--we do not know how quickly, but certainly in less than thirty-four minutes,--the depression, had the spot been at the edge, would have been removed. This change, however, like the existence of the depression, would doubtless not have been discernible by ordinary vision. [Illustration: Fig. 25. Fig. 26. ] Now, let us consider the second observation mentioned above. [Illustration: Fig. 27. Fig. 28. Fig. 29. ] On Thursday, June 26, 1828, the second satellite of Jupiter was about to make a passage across the planet's face. It was observed, just before this passage or transit began, in the position shown in fig. 27 by the late Admiral Smyth. He was using an excellent telescope. It gradually made its entry, looking for a few minutes like a small white mountain on the edge of the planet, and finally disappeared. The reader must understand that the moon was not hiding itself behind the planet, but was on _this_ side of it, and simply lost to view because its brightness was the same, or very nearly the same, as that of the planet's edge. (Its place is shown in fig. 28, but of course the little dark ring was not so seen.) "At least 12 or 13 minutes must have elapsed," says Smyth, "when, accidentally turning to Jupiter again, to my astonishment I perceived the same satellite outside the disc," as shown in fig. 29. It remained visible there for at least four minutes, and then suddenly vanished. To show that the observation was not due to any local or personal peculiarity, it is only necessary to mention that two other astronomers, Mr. Maclear at Biggleswade, and Dr. Pearson at South Kilworth, observed the same extraordinary behaviour of Jupiter's second satellite. The three telescopes are thus described by Admiral Smyth,-- Mr. Maclear's, 3⅓ inches in opening, 3½ feet long; Dr. Pearson's, 6¾ inches in opening, 12 feet long; Adm. Smyth's, 3½ inches in opening, 5 feet long;