It is easily seen by the direction of the arrows that the revolution of the satellites, projected upon the plane of the orbit, is entirely retrograde. All dogmatic affirmations to the contrary are absolutely erroneous. These satellites turn like the hands of a watch,--from left to right, looking at the upper part of the circles. The error of General Drayson's medium comes from the fact that she maintained that the south pole of Uranus was turned toward us at the date of its discovery. Now, in 1781, the system of Uranus occupied relatively to us almost the same situation as in 1862, since the time of its revolution is eighty-four years. It is evident from the figure that, at that moment, the planet presented to us the pole most elevated above the ecliptic; that is, its north pole. General Drayson allowed himself to be led into error when he adopted without verification these paradoxical premises. As a matter of fact, if Uranus had presented to us its south pole in 1781, the movement of the satellites would have been direct. But the observations of the angle of position of the orbits at the time of their passage of the nodes gives us abundant evidence that it was really the north pole which was at that moment turned toward the sun and the earth,--a fact which renders direct movement impossible, retrograde movement certain. [Illustration: Fig. 1--The inclination of the system of Uranus. Aspects seen from the earth at the four extreme positions.] For greater clearness, I have placed outside of the orbit, in Fig 1, the aspect of the system of Uranus seen from the earth at the four principal epochs of the revolution of this distant planet. It is evident that the apparent method of the revolution was analogous to that of the hands of a watch in 1781 and 1862, the opposite in 1818 and 1902. At these dates the apparent orbits of the satellites are almost circles, while during the passage of the nodes, in 1798, 1840, and 1882, they are reduced to straight lines. Figure 1a completes these data by presenting the aspect of the orbits and the method of revolution for all the positions of the planet, even down to our own epoch. I have desired to completely elucidate this question, which is a little technical. _To my great regret_, the spirits have taught us nothing, and this example, to which so much importance is attached, is seen to be an error.[17] Aksakof cites, in this same chapter (p. 343), the discovery of the two satellites of Mars, also made by Drayson through a medium, in 1859; that is to say, 18 years before their discovery, in 1877. This discovery, not having been published at the time, remains doubtful. Furthermore, after Kepler had pointed out its probability, this subject of the two satellites of Mars was several times discussed, notably by Swift and Voltaire (see my _Popular Astronomy_, p. 501). This is not, then, to be set down as an undeniable instance of a discovery made by the spirits.