338 Rev. Mr, Emmett's [Not. . 



call meteoric stones, and are reminded of the illustrious Baron 

 Munchausen's origin of volcanic eruptions. All might fit toge- 

 ther very well, were the data good. That there is an atmosphere 

 appears highly probable, because volcanic eruptions have been 

 frequently observed ; or at least, phaenomena have been seen 

 which cannot be explained upon any other hypothesis ; the late 

 Sir W. Herschel observed three such phaenomena, April 19, 

 1787; also May 4, 1783; a similar, but more remarkable 

 appearance presented itself March 7, 1794; April 13, 1793, and 

 Feb. 5, 1794, Mr. Piazzi observed a similar phaenomenon; since 

 this time, others have been observed, one by Capt. Kater, and 

 one by myself, Jan. 22, 1825, in the same place as those 

 observed by Piazzi, near the spot called Aristarchus by Cassini, 

 and Mons rorphyrites, by Hevelius. 



In reasoning upon inaccessible objects, we must argue from 

 what we know ; and since, on the earth, fire cannot be main- 

 tained without air, we have a right to make the same 

 assumption respecting the moon : hence then these appearances 

 render the existence of an atmosphere highly probable. Mr. 

 Schroeter endeavoured to establish the existence of a lunar 

 atmosphere, by the prolongation of the cusps, when the moon is 

 two or three days old. At an occultation of Jupiter's satelhtes, 

 the third became indistinct, 1^ or 2^ before it disappeared: the 

 fourth became invisible when near the limb. (Phil. Trans. 1792.) 



Some astronomers have found that stars disappeared suddenly 

 behind the moon, without undergoing any previous change ; 

 whilst others observed a sensible diminution of brilliancy. Now 

 if the subject be fairly examined, we shall find that every occul- 

 tation is not suitable for the purpose ; for, first, except near the 

 change, the moon's light is so great that small changes in the 

 brilliancy of a star cannot be easily observed : secondly, the 

 moon's atmosphere must be of small extent, and very rare ; for, 

 according to Dr. Wollaston (Phil. Trans. 1822), were the earth's 

 atmosphere of infinite extent, the density of the atmosphere of 

 each planet would be equal to that of the earth's, at distances 

 which are proportional to the square roots of their quantities of 

 matter : hence in this extreme case, the lunar atmosphere must 

 be very rare ; but its density is far less than even this quantity : 

 thirdly, for both these reasons, if a star pass nearly perpendicu- 

 larly mto the atmosphere ; i. e. if it be very nearly in the path of 

 the moon's centre, the trifling effect that may reasonably be 

 expected cannot continue longer than one or two seconds of 

 time : on the contrary, if the star be very nearly the moon's 

 semidiameter on either side of the path of the centre, the effect 

 of the atmosphere will be chiefly visible in elevating the star, or 

 in keeping it in view for some seconds after it has been seen to 

 touch the limb. In a number of the Annals for the beginning of 

 . the year 1819, I described a very satisfactory observation. On 

 Ma^ 12. 1826. the star 5 s was similarly eclipsed: theconclu- 



