226 



KNOWLEDGE. 



[OCTOBEK 1, 1900. 



not occur near the crossing of any coronal rays, in which 

 situation such an object might possibly be simply an 

 interspace, but appears to encroach on bright rays. 

 Morever, it is the centre of three aa-cs of circles, concave 

 towards the sun, with radii of 3', 6', and 10' respectively, 

 the middle one being fairly strong, while the others 

 are excessively faint. These singular appeai-ances were 

 thought by Mr. Ranyard to indicate the existence of a 

 comet, showing as a dark object on the background of 

 the corona; but as to this I pass no opinion. If it was 

 a comet, its appearance was unique, for the comet on 

 the negatives of the 1882 eclipse, and the much fainter 

 one found by Schaeberle on his photographs of the 

 eclipse of 1893, were both bright objects. But in any 

 case it seems impossible that the dark spot and concen- 

 tric arcs on the corona of 1871 can be interspaces 

 between rays, for the arcs actually cut through 

 several coronal rays almost at right angles, partially 

 obliterating them. The whole appearance is extremely 

 difficult to see, but I have traced the dark spot and the 

 arcs on several negatives of two different series, and am 

 certain of their existence. 



The next example of dark markings occurred in the 

 corona of 1896, the eastern side of which exhibited 



Pro 



Fig. 2. — Diagram of Markings on the Coronn of 1896. 



features of a different kind to any I have examined. I 

 will, however, refer only to those bearing on the subject 

 in hand. Almost at the sun's equator is a bright 

 double-headed prominence, which is distinctly outlined 

 by a strong daa-k line, following all its contours. A 

 little to the north is a small hooked coronal ray about 

 2i' high, apparently springing from a small prominence. 

 This ray is also outlined in the same manner. But 

 by far the most extraordinary appeai'ance is that of a 

 dark roughly elliptical ring, about IV in its longer 

 axis, which stands on the top of the bright prominence. 

 From the summit of the ring springs a fairly bright, 

 fine ray, which would probably be traceable further 

 down towards the limb but that its base seems cut off 

 by the ring. There are many other dark streaks in this 

 part of the corona, but we may confine our attention 

 to the most striking features — the outline to the 

 prominence and the ring. They are clearly seen on at 

 least two of the negatives taken by Mr. Shackleton m 

 Sir G. Baden-Powell's expedition to Nova Zembla, and, 

 unlike the markings on the corona of 1871, they are 

 quite easy to see under suitable conditions of illumina- 

 tion. A little reflection will convince anyone that the 

 outline can be due to no known photographic effect. The 

 image of a bright object (such as a bright prominence) 

 may spread itself on the islate, and thus appear en- 

 larged, or it might conceivably be surrounded by a 

 halation ring, though I feel svu'e that the exposures 

 during eclipses have never been nearly sufficient to cause 



such a ring round a prominence. But neither of these 

 well-known photographic effects will explain the appear- 

 ance in the least. Had an observer drawn the dark 

 outline surrounding the bright prominence, we should 

 have concluded at once that it was a mere effect of 

 contrast, but the camera is fortunately not influenced 

 by conti-ast. Is it possible that the prominence had 

 edges enormously brighter than its centre, so that the 

 dai'k outline is a phenomenon of reversal? This is 

 improbable in the last degree, in view of the small 

 aperture of the instrument and its considerable focal 

 length ; there was also slight hazy cloud, and the plates 

 generally show no signs whatever of over exposure; 

 their definition is admirable. There are many instances 

 of reversal of the images of prominences in 1882, 1893, 

 1898, and 1900. In 1882 their centres were reversed, 

 but there has been no case of reversal of their 

 edges. Besides, the hooked coronal ray is also outlined, 

 and that was certainly not bright enough for reversal, 

 so this explanation breaks down. Then we have the 

 elliptical ring, for which there seems absolutely no ex- 

 planation, except that it is really a dark marking of 

 some kind. It is surely absurd to suggest that it can 

 be a mere space between coronal rays ; we should have 

 to imagine a tunnel cut through the body of the corona, 

 directed precisely in the line of sight, and a plug of 

 coronal matter lying along the centre of that tunnel but 

 not touching its sides. 



It will be seen that in the above cases the argument 

 for the objective existence of dark markings is based 

 upon the form and character of the markings, and 

 not upon their actual darkness. Neither in 1871 nor 

 in 1896 are they nearly as dark as the sky; but have 

 we any instances of markings in the corona that are 

 actually darker than the sky? If so, it appears to me 

 that their objective existence is jDroved beyond a doubt. 

 I believe we have such evidence, but here great caution 

 is required, for although the camera is, as has been said, 

 unaffected by contrast, the eye which examines the 

 photographs is much affected by it, and we may be very 

 easily deceived. 



I have before me two negatives of the eclipse of 1898, 

 taken by Mr. F. Bacon at Buxar, near Benares ; they 

 are rather over-developed, the lower joortions of the 



Fig. 3. — Diagram of Markings on tlie Corona of 1898. 



corona are extremely dense and opaque, but the focus 

 is excellent, and the outer portions well shown. The 

 scale is a little over half an inch for the moon's 

 diameter. On first looking at these I was struck by 

 the unusual sharpness of definition of some of the rays 

 of the great southern rift; but on more careful examinar 

 tion with various illuminations it seemed possible that 

 this sharp definition is due to nothing less than to two 

 or more fine dark rays (of course bright on the negative) 

 lying between some of the bright polar rays near the 

 western boundary of the polar rift. There is nothing un- 



