TOTAL ECLIPSE OF THE SUIST, APRIL 16, 1893. 
577 
The visible crescent of photosphere when these three photographs were taken was 
in the south-west quadrant, and as the dispersion was in the north and south 
direction, the cusp nearest to the south point was the one most favourably situated 
for showing bright arcs. The light of the photosphere is drawn out into a continuous 
hand hy the action of the prism, and this band is of varying intensity in consequence 
of the irregularities of the moon’s limb, the so-called Baily’s beads being drawn out 
into longitudinal streaks, as was seen in a photograph taken out of totality in 1882. 
In Photograph No. 22 this photospheric spectrum is crossed by bright arcs which 
also project some distance beyond in both directions, and no Fraunhofer lines are 
visible ; the bright arcs corresponding to H^, H y , H s , H, and K, and other ultra¬ 
violet lines of hydrogen are especially intense. The group of prominences near the 
south pole is also depicted in these radiations, as well as in H a and D 3 . A very large 
number of short bright arcs is also seen at the cusp, but these cannot be traced 
across the spectrum of the photosphere. 
An attempt to indicate the appearances of different typical arcs is made in fig. 7, 
as the detail is too minute to be satisfactorily shown in a photographic reproduction. 
The brighter ones show great irregularities throughout their lengths, and the fainter 
Appearances of bright arcs in spectrum of cusp. 
ones appear only in the parts corresponding to the brightest regions. None of the 
arcs are brightest close to the edge of the continuous spectrum, and the faintest 
of them have the appearance of being detached as in the case of Fe 4072 in the 
diagram. In the Photograph No. 22 the spectrum of the corona is also seen ; as a 
faint continuous one at the eastern and western edges on the illuminated sky back¬ 
ground, and an arc corresponding to 1474 K is visible in the south-western part. 
A portion of this photograph is given in Plate 13. It cannot be satisfactorily 
reproduced in its entirety owing to the great density of the negative. The general 
appearances of Photographs 23 and 24 (see Plate 13) are very similar to that of 
No. 22, but there are several important differences in detail. The photospheric 
MDCCCXCVI.—A. 4 E 
