(iii) For 4-i minutes, starting at the count " Thirty " after the beginning of totality and ending at the count " Sixty, five minutes gone."
(iv) For 2 seconds, when I gave the signal " Now " at the end of totality.
Lieutenant Briggs carried out the programme with admirable precision.
Result. All four photographs are of value. The first serves for determination of the scale of the spectrum and shows that the focussing was very successful. The second and fourth photographs show both bright and dark lines, as was intended ; the second is a weak negative, but is full of interesting details ; the fourth is quite strong, teeming with bright crescents, of which more than ninety have been counted between A 4308 and A 4405. Many of the crescents are bright through- out their whole length ; many are reversed in the middle ; some of the bright crescents are accompaned by dark crescents on either side ; others have a dark crescent only on one side or the other ; all varieties seem to be represented.
The third photograph is not very strong but it shows marked con- tinuous spectrum, with the blue coronal ring distinctly though not easily visible. The edges are ill-defined, in marked contrast with the sharp prominence-tips that serve to mark out the positions of the H y and Hs chromospheric rings.
The radial extension of the blue coronal ring is about 3', but there is no sign of radial structure.
Two determinations of the wave-length of the light forming the blue ring give the values 4231 '4 and 4231 '9, if the wave-lengths ascribed to the dark H y and HS lines may be adopted for the bright hydrogen prominences.
The dispersion is about 5' 14 tenth-metres to a millimetre, and the scale of the ring is such that its diameter corresponds to 65 tenth- metres. The conditions are thus favourable for the detection of coronal rings. No ring except that at 4231 has been with certainty detected between H v and H$.
17. Visual Objective-grating Spectroscope.
An objective grating spectroscope was used as in India, in 1898,* for visual observations of the green coronal ring, with a view to (i) repeating the search made during the Indian eclipse for fine radial streamers, (ii) comparing the distribution of light in the green ring with that in the blue ring near A 4231, of which I hoped to get a photograph with a photographic objective grating camera.
A plane grating by Eowland, 14,438 lines to the inch, on a ruled
- ' Roy. Soc. Proc.,' vol. 64, p. 57; ' Monthly Notices, Eoy. Astro. Soc.,' vol. 58,
App., p. [57].
