June 24. 1897] 



NA TURE 



175 



THE APPROACHING TOTAL ECLIPSE OF 



THE SUN} 



II. 



THE considerations which led me, in 187 1, to employ a 

 spectroscope without collimator may here again be 

 summarised. If in an ordinary spectroscope, the straight 

 slit be replaced by a circular one, bright rings replace the 

 bright lines which are ordinarily seen in radiation spectra, 

 and since in the solar surroundings we have chiefly to 

 deal with radiation phenomena, the chromosphere and 

 < orona themselves can be used during an eclipse as ring 

 slits, and on account of their distance, a collimating lens 

 can be dispensed with. 



In the report on the eclipse of 1875, by Dr. Schuster 

 and myself, the principles of the method, as applying to 

 photographs taken during totality, were stated as follows 

 /'////. Trans., 1878, Part i, p. 139) : — 



" Supposing that the corona and chromosphere only 

 send out the same homogeneous light, one image only 

 will appear on the sensitive plate, the only effect of the 

 prism being to displace the image. As far as the pro- 

 tuberances are concerned we know they give a spectrum 

 of bright lines, and we expect, therefore, to find on the 

 plate each protuberance represented as many times as 

 it contains lines in the photographic region. The different 

 protuberances would be arranged in a circle round the 

 sun, and these circles would overlap or not, according 

 to the dispersive power of the prism and the difference 

 in refrangibility of the lines. ... If the corona gives 

 a series of bright lines we shall find a series of outlines 

 on the photographs similar to that corresponding to the 

 protuberances. . . . If we find that part of the corona 

 .;ives a continuous spectrum, that part alone will be 

 Irawn out into a band." 



To this it may be added, that successive photographs 

 will differ on account of the difference of phase. One 

 part of the chromosphere will be visible at the beginning 

 of totality, and another part at the end. The smaller 

 prominences visible at the beginning of totality are sub- 

 sequently eclipsed by the moon, and their spectra are 

 consequently absent from later photographs, while a new 

 prominence region makes its appearance. In the same 

 way, the part of the corona the spectrum of which is 

 photographed will vary at different phases, but only in 

 the lower parts. 



The results obtained by Prof. Respighi and myself 

 during the eclipse of 1871 in India, in which part of the 

 attack consisted in the employment of slitless spectro- 

 scopes—a method of work at which we had arrived 

 independently — indicated the extreme value of such 

 observations. 



For my own observations in 1871 I had arranged a 

 train of five prisms without either collimator or observing 

 telescope. " I saw four rings with projections defining 

 the prominences. In brightness, C came first, then F, 

 then G, and last of all 1474K.. Further, the rings were 

 nearly all the same thickness, certainly not more than 

 2' high, and they were all enveloped in a band of 

 continuous spectrum" (Nature, vol. v. p. 218, 1872). 



Respighi's observations were made with a telescope of 

 a\ inches aperture, with a large prism of small angle in 

 front of the object-glass. The principal results obtained 

 by him were as follows (Nature, vol. v. p. 237, 1872) : — 



"At the very instant of totality, the field of the tele- 

 scope exhibited a most astonishing spectacle. The 

 chromosphere at the edge, which was the last to be 

 eclipsed, . . . was reproduced in the four spectral lines, 

 C, D-j, F and G, with extraordinary intensity of light. . . . 



" Meanwhile the coloured zones of the corona became 

 continually more strongly marked, one in the red corre- 

 sponding with the line C, another in the green, probably 

 coinciding with the line 1474 of KirchhofPs scale, and 

 a third in the blue perhaps coinciding with F." 



J Continued from page 157. 



NO. 1443- ^'^L. 50] 



" The green zone surrounding the disc of the moon 

 was the brightest, the most uniform and the best 

 defined." 



My observation (Brit. Assoc. Report, 1872, p. 331) was 

 made intermediately between the two observations of 

 Prof. Respighi. The observations may be thus com- 

 pared : — 



Respighi C D3 ... F.G. Chromosphere and prom- 



inences at beginnmg of 

 totality. 



Lockyer C 1474 (faint) F.G. Corona 80 sees, after be- 

 ginning of totality. 



Respighi C I474 (strong) F. Later. 



I had no object-glass to collect light, but I had more 

 prisms to disperse it, so that with me the rings were 

 not so high as those observed by Respighi, because I 

 had not so much light to work with ; but such as they 

 were, I saw them better, because the continuous spec- 

 trum was more dispersed, and the rings (the images of 

 the corona) therefore did not overlap. Hence, doubt- 

 less Respighi missed the violet ring which I saw ; but 

 both that and 1474 were very dim, while C shone out 

 with marvellous brilliancy, and D3 was absent. 



In arranging for the eclipse of 1875 '" ^iam and the 

 Nicobars, the method was further developed by the in- 

 troduction of photography, and the first results of this 

 extension were given in the Report of the Eclipse Ex- 

 pedition of that year. They showed clearly that with 

 the rapid dry plates of to-day a considerable increase 

 of dispersion might be attempted. 



The object-glass employed on this occasion had an 

 aperture of 3I inches and a focal length of 5 feet, while 

 the prism had a refracting angle of 8 degrees. 



Two photographs were obtained with exposures of one 

 and two minutes respectively. Both are reproduced in 

 the Report {Phil. Trans., 1878, vol. clxi.x. Part i. p. 139), 

 and they show only such differences as can be attributed 

 to difference of phase. The dispersion was very small 

 compared with the size of the sun's image, so that the 

 photographs present the appearance of an ordinary photo- 

 graph of the eclipsed sun, which is slightly distended 

 in the direction of dispersion. The various prominences 

 each show three images, two of which were identified 

 with H/3, Hy, while the others were found to correspond 

 to a wave-length of about 3957. 



It was suggested (Report, p. 149) that this repre- 

 sented the H and K radiations of calcium, and this is 

 fully confirmed by the results obtained in 1893, to say 

 nothing of results obtained in other eclipses. 



I next proceed to remark very briefly upon the photo- 

 graphic results obtained since 1875. In 1878, near the 

 sun-spot minimum, the method was employed by several 

 observers, myself among them, but no bright rings w^ere 

 recorded. The maximum sun-spot conditions previously 

 observed had entirely changed ; indeed with a slit spec- 

 troscope the 1474 line was very feeble, and was only seen 

 by a few of the observers, and hydrogen lines were 

 similarly feeble {American Journal of Science, vol xvi. 

 P- 243). 



Part of my own equipment for this eclipse consisted of 

 a small grating placed in front of an ordinary portrait 

 camera, and with this I obtained a photograph showing 

 only a very distinct continuous spectrum.^ 



The method was employed by Dr. Schuster in Egypt 

 in 1882 ; the camera was of 3 inches aperture and 20 

 inches focal length, with a prism having a refracting 

 angle of 60° {Phil. Trans., vol. clxxv., 1884, p. 262). 

 The single photograph obtained (not reproduced in the 

 Report) was stated to show two rings, which were con- 

 sidered to be due to the lower parts of the corona, and 

 therefore to correspond to true coronal light. The wave- 



1 With a duplicate grating I observed the spectrum of the eclipsed sun, 

 and again in three different orders, saw nothing but continuous spectrum 

 (Nature, vol. xviii., 1878, p. 459). 



