108 



KNOWLEDGE. 



[May 2, 1898. 



for the special work which we had in contemplation, but 

 it eeemed to us that we might make good use of it. It was 

 fitted with one of Dallmeyer's telephotographic lenses, and 

 had a full aperture of nearly two and a half inches and an 

 equivalent focal length of almost eight feet. The mag- 

 nifying power employed was therefore very nearly twelve 

 diameters. This appeared to us much too high, but the 

 definition which we actually obtained justified Mr. Waters 

 in his choice of so great a scale, and forms the highest 

 possible testimonial to the quality of Mr. Dallmeyer's 

 optical work. Our idea was that with the two cameras 

 we might obtain a series of photographs, the equivalent 

 exposures of which might be arranged so as to form a geo- 

 metrical series ranging from the shortest up to Taylor's 



over this one hundred and forty degrees the "coronium" 

 ring was perfectly continuous, it was interrupted by no 

 rifts, it spread out into no rays. It was traceable to a 

 height of between five and six minutes from the moon's limb, 

 and corresponded therefore very closely to what to the eye 

 appeared to be the brightest inner corona. So far, then, 

 "coronium " appears to be pretty evenly distributed round 

 the sun, and not to follow the striking and characteristic 

 forms which attract such notice in the corona as seen 

 directly. 



Our second field of work, the attempt to photograph the 

 long coronal streamers, met with a most wonderful success. 

 Our two long-exposure plates — four times Taylor's Umit — 

 equivalent to one hundred and twenty seconds with F/15 — 



Sun's Equator. 



limit. The common ratio in this case was about four and 

 a half, and, taking Taylor's limit as unity, ranged from one 

 four-hundredth up to one. We thought this additional 

 item would not be devoid of interest, as it seemed to us 

 that not a few coronal photographs on former occasions 

 had failed to be so successful as they might have been 

 through over exposure. 



So much as to our programme ; now as to our results. 



With the prismatic opera-glass, my first sensation, after 

 watching the indescribably beautiful changes which the 

 spectrum underwent as totality came on, was one of pro- 

 found disappointment. The continuous spectrum entirely 

 swamped nearly all the bright rings except those of 

 hydrogen and hehum ; and the 1474 K ring, either because 

 it was fainter on the eastern side than on the western, or 

 because my eyes took some time to become attuned to the 

 light, was not seen till after the middle of totality. It was 

 then detected over an arc of about one hundred and forty 

 degrees — that is to say, only in the semicircle round the 

 point of third contact, and not over the whole of that, as 

 the continuous spectrum was necessarily so bright at its 

 two edges as to swamp the bright line spectrum there. Still, 



showed the four principal coronal streamers to a much 

 greater distance than ever before. They can be clearly 

 and unmistakably followed to distances from the moon's 

 centi'e of three, four, four and a half, and six diameters 

 respectively. In the case of the great south-west ray we 

 believe that we have traced it, though very feebly, under 

 suitable illumination, to a much greater distance still, and 

 we may have further remarks to make upon its details on 

 some future occasion. As it is, however, it is beyond 

 challenge that this south-west ray is shown clearly and 

 distinctly upon two photographs much further than has 

 been the case in any photograph of the corona that has 

 ever been taken before. 



Indeed on three photographs. For beside these two long- 

 exposure photographs taken on " S an dell " plates (kindly 

 developed by Mr. J. T. SandeU himself), a third photo- 

 graph taken with one quarter the exposure — practically 

 Taylor's limit — an " Ilford extra rapid " plate which we 

 developed ourselves, might have claimed, had it not been 

 for the two " SandeU '' plates, to have given the record 

 coronal extension. 



The orientation of these three plates is given with great 



