TOTAL ECLIPSE OF THE SUN, APRIL 16, 1893. 
597 
-explained when certain plates are employed. A positive photograph of the corona 
of 1893, some 6' high, illuminated by a large gas flame or electric lamp, forms an 
artificial eclipse, which has been viewed with instruments of different dispersions, at 
a distance such that the diameter subtended an angle of about 32'. With a very 
small dispersion, similar to that employed by Respighi in 1871 in relation to the 
apparent size of the ring, one sees circular spectra, approximating very closely to 
those described and figured by Respighi the effect is best seen when the green 
ring is intensified by introducing a salt of thallium between the poles of the electric 
arc, when the latter is employed as a source of light. A green ring, sharply defined 
on its inner edge, then becomes visible, a red ring is very distinct hut ill-defined, and 
a blue ring is still less distinct. With increased dispersion, similar to that I employed 
in 1871, all traces of the blue and red rings disappear, and the green one is only 
visible so long as there is a special green radiation. 
These experiments suggest that the very broad blue and red rings recorded by 
Respighi in 1871 are possibly not conclusive evidence that the H a and radiations 
of hydrogen were very strong in the spectrum of the corona, as rings strikingly 
resembling them can be produced by a continuous spectrum alone when viewed with 
similar instrumental conditions. At the same time no such doubt is thrown upon 
my own observations of narrow rings which were made with the dispersion of five 
prisms, as false rings of red and blue light are not seen under these conditions. The 
red and blue hydrogen rings, as I saw them in 1871, were about 2' high, and 
bright compared with 1474 K; my observations were made earlier in the eclipse 
than those of Respighi, to which reference is made above, but the fact that I did not 
see D 3 seems conclusive evidence that I was not observing rings due to the chromo¬ 
sphere and prominences. Hydrogen was, therefore, certainly present in the corona 
of 1871 to a height of about 2', but the observations suggesting its spectroscopic 
visibility at a greater height are not conclusive. 
With regard to the photograph of 1875, in which H y was definitely ascribed by 
Dr. Schuster and myself to the corona, the absence of corona rings corresponding to 
H and K, which are the strongest lines of the chromosphere and prominences, appears 
to support the idea that the H y ring was not due to chromosphere or prominences. 
A photograph of the artificial corona, illuminated by an incandescent lamp, under 
instrumental conditions closely approximating to those employed in 1875, shows only 
a liazy ring near H v , corresponding to the region of maximum sensitiveness of the 
plate ; in this way a sharply defined ring similar to that on the photograph cannot 
be obtained. 
It is probable, therefore, that in 1875, as well as in 1871, hydrogen was present in 
the inner corona. 
My observations in 1882f also gave indications of hydrogen. 
* ‘ Atti della Reale Accad. dei Lincei,’ March 3, 1872, p. 17. 
t ‘ Nature,’ vol. 26, p. 101. 
