232 The Eclipse of Last December. [April, 



obvious that the air above and around the observer of a total 

 eclipse must be lit up by the prominences and sierra, since 

 these are actually seen ; and, a fortiori, it must be lit up by 

 that intensely bright inner portion of the corona which has 

 so often led the observers of total eclipses to suppose that a 

 ring of real sunlight has, in fact, remained unconcealed ; 

 though it must be noted at the same time, as I pointed out 

 in the paper just quoted, that the light from this source 

 "would extend over the moon's disc, since it would illumi- 

 nate the air between the observer and the moon's body." 

 This, indeed, as we shall presently see, has been abundantly 

 confirmed by the observations made during the recent 

 eclipse. 



It is clear that, even though the corona thus accepted 

 as real by the supporters of the glare theory, should, in 

 fact, be but a portion of the complete corona, the enuncia- 

 tion of the theory could in no way interfere with the ob- 

 servations to be made by the eclipse expeditions. On the 

 contrary, the very attempt to distinguish where the real 

 corona ended and the atmospheric corona began, could 

 scarcely fail to be rewarded by results of extreme importance. 



Now, so far as the evidence available before the late 

 eclipse was concerned, we may select direct observations 

 formerly made, and the spectroscopic observations made 

 during the Indian and American eclipses (of 1868 and 1869) 

 for consideration. Photographic evidence was not wholly 

 wanting, but so far it was certainly unsatisfactory. Polariscopic 

 evidence, again, had seemed (and still seems) most vague 

 and contradictory. 



But direct observation had already revealed many facts 

 bearing most importantly on the question of the corona's 

 real extent. 



To begin with, on the very point which was assigned 

 strangely enough by the organising committee as the 

 principal object of the expeditions, " the differentiation of 

 the outer layers of irregular outline, and a stratum, say 

 some 5' or 6' high close around the sun," the history 

 of coronal research is decisive. Plantade and Capies, in 

 1706, had already detected the difference, which 164 years 

 later has been called " the key-note of all the observations " 

 made last December. In 1842, Arago had dwelt specially 

 on the same point. In fact, nothing in astronomy has been 

 more thoroughly determined than the general uniformity of 

 the coronal light up to a certain moderate distance from the 

 sun, where a rapid but not sudden degradation takes 

 place. 



