Solar Eclipse of July, 1860. 379 



outlines of the lunar disc ? These flames were either in the moon 

 or in the sun, or in our atmosphere ; unless, indeed, denying their 

 actual existence, we regard them as an effect of light, for example 

 as phenomena of diffraction. 



The two last snppositions have found few partizans. Before 

 adopting any hypothesis it is necessary to decide by observation a 

 certain feature of the phenomenon. During the eclipse, the disc 

 of the moon moves across the disc of the sun. But do these red- 

 dish clouds follow the moon in its movement ? or does each cloud 

 remain invariably above the same point on the solar disc? In 

 the first case the origin of the luminous clouds is to be sought in 

 the moou ; in the second case, these clouds belong to the sun. 

 For clearness sake, assume the latter supposition, and observe 

 what appearances should present themselves when the lunar disc 

 passes like a screen over the whole. Consider first a cloud situ- 

 ated on the east and adherent to the sun's limb. This object will 

 be visible at the instant when totality commences. The advanc- 

 ing moon will regularly, at the rate of a half second of arc in a 

 second of time, cover with its imb successively the lower, then the 

 middle and lastly the higher portions — thus constantly diminish- 

 ing the height of the cloud. For a cloud situated on the west 

 these appearances will be reversed, its magnitude increasing as 

 the moon gradually uncovers it. If then the roseate appendages 

 seen during a total eclipse, depend on the sun, the fact should ap- 

 pear by the variation in height between those which appear in the 

 east and the west. The phenomena will appear otherwise if the 

 clouds appertain to the moon. 



In the absence of equatorial solar clouds, the question in dispute 

 can still be decided by observations on those seen on the south or 

 north of the disc. The height of these clouds ought not to vary 

 it is true, whether they belong to the moon or to the sun, but in 

 the latter case carried away by the sun they would be displaced 

 in the lunar disc with a certain velocity, while if they are adhe- 

 rent to the moon's disc, they would not be displaced. Hence 

 the study of the height of the luminous clouds, whether east and 

 west or north and south, has the highest interest. All the ele- 

 ments of the desired demonstration are found in the Spanish ob- 

 servations. In my first report, I mentioned the successive increase 

 in thickness of a band of rosy clouds visible from the east to the 

 end of the eclipse. Messrs. Yvon Villarceau and Chacornac have 

 carefully noted the motions of a cloud, situated on the north 



