NATURAL PHILOSOPHY. 183 







barometer and thermometer were observed at two heights ascertained 

 by observation, one considerably above the other, and both above the 

 region in which the currents from the equator influence the tempera- 

 ture, data would be furnished by which an approximate determination 

 of the height of the atmosphere might be attempted. 



AUGMENTATION OF THE APPARENT DIAMETER OF A BODY BY 

 ITS ATMOSPHERIC REFRACTION. BY PROF. CHALLIS. 



For reasons given in the preceding communication, it was assumed 

 that atmospheres generally have definite boundaries at which their 

 densities have small but finite values. Two cases of refraction were con- 

 sidered : in the one, the curvature of the course of a ray through the 

 atmosphere was assumed to be always less than that of the globe 

 it surrounds ; and in the other, the curvature of the globe might be 

 the greater. The former is known to be the case of the earth's atmos- 

 phere ; and it was supposed that, a fortiori, this must be the case with 

 respect to any atmosphere the moon may be supposed to have. On 

 this supposition it was shown that the apparent diameter of the moon, 

 as ascertained by measurement, would be greater than that inferred 

 from the observation of an occultation of a star, because, by reason 

 of the refraction of the atmosphere, the star would disappear and 

 reappear when the line of vision was within the moon's apparent 

 boundary. The same result would be obtained from a solar eclipse. 

 It was stated that, by actual comparisons of the two kinds of determi- 

 nations, such an excess to the amount of six to eight seconds was 

 found. This difference may reasonably be attributed to the existence 

 of a lunar atmosphere of very small magnitude and density. The 

 author also stated that from this result there would be reason to 

 expect, in a solar eclipse, that a slender band of the sun's disc imme- 

 diately contiguous to the moon's border would be somewhat brighter 

 than the other parts, and advised that especial attention should be 

 directed to this point on the next occurrence of a solar eclipse. The 

 case in which the curvature of the path of the ray is greater than 

 that of the globe was assumed to be that of the sun's atmosphere ; 

 and it was shown, on this supposition, that all objects seen by the 

 rays which come from the sun's periphery are brought by the refrac- 

 tion to the level of the boundary of the atmosphere, whether they pro- 

 ceeded from objects on the surface of the interior globe, or from clouds 

 supposed to be suspended in the atmosphere. Accordingly, the con- 

 tour of the sun should appear quite continuous, and the augmentation 

 of apparent semi-diameter will be equal to the angle subtended at 

 the earth by the whole height of the atmosphere. The apparent 

 diameters of the planets will, for like reasons, be augmented to a cer- 

 tain amount by the effect of refraction ; and, on account of the great 

 distances of these bodies from the earth, the eclipse of a satellite will 

 take place as soon as the visual ray is bent by the interposition of the 

 atmosphere. 



ATMOSPHERIC WAVES. 



One of the most important results which have flowed from recent 

 systematized meteorological observations has been the determination 

 12 



