HISTORY OF PHYSICAL ASTRONOMY.

231

to afford similar indications of a lunar atmosphere without placing thequestion of its existence beyond all doubt. When the star comes upto the moon, it is observed frequently to hang on the margin of thedisk for several seconds previous to its disappearance. An effect ana-logous to this would manifestly be produced by the interposition of thedenser strata of the moon's atmosphere between the star and the observer.On the occasion of the first of the two observations of M. Arago, abovecited, the star appeared to adhere to the moon’s limb for the space of threeor four seconds after it was perceived to be in contact with it. Thiscircumstance could not be ascribed to the effect of irradiation, since theoccultation took place at the dark limb of the moon, which was visible bymeans of the lumiere cendree , or reflected light of the earth. Numeroussimilar instances of the star’s adherence to the limb of the moon havebeen recorded by astronomers, but, as in the case of a diminution of lustre,the phenomenon is not general. In the vast majority of occultations thestar comes up to the limb of the moon, and then disappears instantaneously.Its occasional adherence to the limb cannot be admitted, therefore, to af-ford any conclusive proof in favour of a lunar atmosphere. There is a thirdcircumstance connected with the phenomena of occultations which is cal-culated to throw light upon this interesting question. It is manifest thatif the moon be surrounded by an atmosphere, the refractive power of thelatter would cause a star to be visible during a short time subsequent toits immersion, and also during an equal lapse of time previous to itsemergence. The interval, therefore, which elapses between the disappear-ance of the star at the moon’s limb, and its subsequent re-appearance atthe opposite limb, ought to be less than the time which the moon takes todescribe an arc equal to her apparent diameter, by twice the horizontal re-fraction of the lunar atmosphere. When the interval, however, is calcu-lated by means of the theory of the moon’s motion and her apparent dia-meter, corresponding to the time of occultation, the result is found to ex-hibit an exact accordance with that derived from direct observation. Itfollows, therefore, that if the moon really be surrounded by an atmosphere,its refractive power must be very inconsiderable.

The occultation of a planet by the moon suggests a peculiar mode oftesting the existence of a lunar atmosphere, in addition to thosealready alluded to, which are also applicable to the fixed stars. Whenthe planet has so nearly approached the moon as to be almost incontact with her limb, the rays issuing from the margin of its disk willmanifestly undergo different degrees of refraction in the course of theirpassage through the lunar atmosphere, and an apparent distortion of itsfigure will be the necessary consequence. Observation, however, does notafford any indications of such a phenomenon. In 1679 Cassini and LaHire witnessed an occultation of Jupiter by the moon, and they found thatfrom the time of the*planet’s approach to the limb until its complete dis-appearance it did not exhibit the slightest indication of a change of figure*.Similar phenomena have been observed by modern astronomers, but on nooccasion has any distortion of the round appearance of the planet beendistinctly discernible.

The phenomena of solar eclipses have also been employed by astrono-mers for the purpose of establishing the existence of a lunar atmosphere.When the dark body of the moon is projected upon the luminous disk of

* Anc. Mem. Acad, des Sciences, tome i., p 303.