Existence of an Atmosphere surrounding the Sun. 445 



near the circumference, tend so strongly to force upon us the 

 conviction of the existence of an atmosphere of no inconsiderable 

 amount, that it is of some importance to inquire whether the 

 motions of the spots, in passing over the disc, present any 

 favourable or unfavourable indications. 



The inquiry presents difficulties of two sorts, arising, first, 

 from the little suitability of solar spots in general as data in any 

 matter of dehcacy, from their constant liability to change of 

 form; and secondly, from our entire ignorance of the law of 

 density, which should be assumed for the supposed atmosphere 

 under the peculiar conditions of temperature to which it must 

 be subject. The first of these difficulties we must surmount as well 

 as we can by selection of the spots for discussion : the second, I 

 shall make free to eviade on the present occasion by the old 

 assumption of homogeneity, proposing only to inquire what sort 

 of result will be yielded by that hypothesis, which in the case of 

 the earth's atmosphere is well known to afford a close represen- 

 tation of the observed actual effects on a ray of light. 



Let P Ap be a central section of the sun's actual photosphere 

 by a plane passing through E, the centre of the earth. Let 

 II C r be the corresjjonding section of the atmosphere, supposed 

 homogeneous. Let SP = r; Sll = r (l+h); refractive index 

 = m. If a ray of light from the sun's actual limb reach the 

 earth by the path P R E, it will appear to have come from the 

 limb of a concentric sphere (the apparent photosphere), whose 

 section is Q B q, and whose radius Q S = 7W . P S ; since 

 sin Q R S = m . sin P R S. A spot at p on the actual photosphere 

 will appear in the direction E r q, or will be seen at q on the 

 apparent ])hotosphere, under an apparent angle B8 q from the 

 centre of the disc, instead of A S;j the true. The position-angles 

 of spots will be unaffected. 



