230 The Eclipse of Last December. [April, 



relations of tridimensional space, that when the sun is 

 directly behind the moon, as in a total eclipse, his rays 

 passing grazingly by the moon, must so illuminate our 

 own atmosphere as to exhibit a glory of light — a real 

 radiated corona — around the dark disc of the moon ; and it 

 was this consideration — plausible, but beyond question erro- 

 neous — which was the original basis of Faye's " atmospheric 

 glare " theory. Few but mathematicians would have been 

 prepared to hear — as Baxendell, Curtis, Harkness, and 

 others at once pointed out — that the illumination by direct 

 solar light could not in any ordinary total eclipse cause 

 a ray of light to come from a space of many degrees in 

 width all round the place of the eclipsed sun. A negative 

 corona, that is, a corona of blackness softening off all round 

 towards a background of considerable brightness, would be 

 the real appearance resulting from the effect of direct 

 sunlight.* 



This having been proved, the atmospheric glare theory 

 was modified by the introduction of a supposed action 

 exerted by the moon on the solar rays. The nature of this 

 action was not definitely (or indeed at all) stated ; but the 

 theory thus modified agreed, so far as it went, with La 

 Hire's, which had been long regarded as untenable. As 

 a matter of fact, also, while introducing fresh difficulties, it 

 removed none of those urged by Baxendell and others. 



It is bare justice, however, to the supporters of the atmo- 

 spheric glare theory to point out that, as the eclipse of last 

 December approached, the objectionable features of the 

 theory were removed, and views were put forward which 

 undoubtedly agreed much better with observed facts. Let 

 me not be misunderstood, as respects my use of the word 

 objectionable. To an erroneous theory, in the abstract, no 

 objections need be made ; since the enunciation and demon- 

 stration of the true theory disposes of the incorrect one 

 without occasion for a special refutation. But it sometimes 

 happens that the enunciation of an erroneous theory at 

 a particular time may interfere with the progress of obser- 

 vation — either by distracting the attention of observers, or 

 by causing them to under-estimate the interest of the object 

 which they are to study. And beyond question, if the corona 

 might by any possibility be no more noble a phenomenon, 



* On this subject, Sir John Herschel writes to me — "Placing the limits of 

 what may properly be called the earth's atmosphere at 100 miles high, then at 

 the borders of a circular section subtending an angle of about 23 degrees in 

 diameter every trace of light reflected in our atmosphere would be quite 

 evanescent." This precisely accords with the objections I urged last year 

 against the atmospheric glare theory as originally propounded. 



