The total lunar eclipse which occurred on October 4, 1884, was remarkable for the almost total disappearance of the moon during totality. One observer says that "in the open air, if one had not known exactly where to look for it, one might have searched for some time without discovering it. I speak of course of the naked eye appearance."[87] On the other hand the same observer, speaking of the total eclipse of the moon on August 23, 1877, which was a bright one, says-- "The moon even in the middle of the total phase was a conspicuous object in the sky, and the ruddy colour was well marked. In the very middle of the eclipse the degree of illumination was as nearly as possible equal all round the edge of the moon, the central parts being darker than those near the edge." In Roger de Hovedin's _Chronicle_ (A.D. 756) an account is given of the occultation of "a bright star," by the moon during a total eclipse. This is confirmed by Simeon of Durham, who also dates the eclipse A.D. 756. This is, however, a mistake, the eclipse having occurred on the evening of November 23, A.D. 755. Calvisius supposed that the occulted "star" might have been Aldebaran. Pingré, however, showed that this was impossible, and Struyck, in 1740, showed that the planet Jupiter was the "star" referred to by the early observer. Further calculations by Hind (1885) show conclusively that Struyck was quite correct, and that the phenomenon described in the old chronicles was the occultation of Jupiter by a totally eclipsed moon--a rather unique phenomenon.[88] An occultation of Mars by the moon is recorded by the Chinese, on February 14, B.C. 69, and one of Venus, on March 30, A.D. 361. These have also been verified by Hind, and his calculations show the accuracy of these old Chinese records. It has been suggested that the moon may possibly have a satellite revolving round it, as the moon itself revolves round the earth. This would, of course, form an object of great interest. During the total lunar eclipses of March 10 and September 3, 1895, a careful photographic search was made by Prof. Barnard for a possible lunar satellite. The eclipse of March 10 was not very suitable for the purpose owing to a hazy sky, but that of September 3 was "entirely satisfactory," as the sky was very clear, and the duration of totality was very long. On the latter occasion "six splendid" photographs were obtained of the total phase with a 6-inch Willard lens. The result was that none of these photographs "show anything which might be taken for a lunar satellite," at least any satellite as bright as the 10th or 12th magnitude. It is, of course, just possible that the supposed satellite might have been behind the moon during the totality. With reference to the attraction between the earth and moon, Sir Oliver Lodge says--