574 
MR, J. NORMAN LOCK5TER ON THE 
forms of the prominences themselves are very clearly depicted, and it will be seen 
that the scale is sufficiently large to reduce confusion of images to a minimum. The 
general appearances on the photograph correspond very closely with those seen at 
the beginning of totality by Respighi in 187 L, and by Ranyard in 1878. 
Negative No. 8, taken 2 seconds later, is hardly distinguishable from No. 7. In 
both plates the corona is represented chiefly by what appears to be continuous 
spectrum. As was to be expected from the projection consequent on the direction of 
dispersion, this continuous spectrum is brightest at the eastern and western limbs, 
and it is brighter on the eastern than the western side, this region of greater 
intensity corresponding to a lower part of the corona. Besides the continuous 
spectrum, however, there are a few feeble arcs, not seen in the reproduction, 
representing true coronal radiations; these are quite distinct from the chromospheric 
arcs. 
Photographs about Mid-eclipse. 
The arc of chromosphere, seen in Photographs Nos. 7 and 8, was covered by the 
moon when the next plate was exposed, but the associated prominences were not 
completely extinguished by the advancing moon until No. 17 was exposed. The 
upper part of the chromosphere in the south-western quadrant did not appear until 
No. 20 was exposed. Hence, Photographs 9-19 inclusive show no chromospheric 
spectrum, but they give a record of the spectra of the upper parts of numerous 
prominences, as well as of the spectrum of the corona, 
A reproduction of Negative 17, taken on an isochromatic plate, is given in Plate 12. 
It will be seen that the spectra of the prominences are similar to those in No. 7, but 
they are simpler for the reason that the lower reaches were covered by the moon. 
At the extreme right in the photograph there is a feeble image of the bright group 
of prominences produced by the C radiation of hydrogen ; as the plates are scarcely 
sensitive at all to the red, the image must in reality have been very intense; 
I) 3 images of the prominences are also seen in the photographs. 
The spectrum of the corona in Negatives 9-19 is to a large extent continuous, but, 
in addition, it is represented in some of the photographs by a nearly complete ring 
corresponding to the 1474 K line, and smaller portions of fainter rings. 
The continuous spectrum is brightest near the photosphere, and fades out very 
gradually at heights depending upon the exposure and development of the plates 
and the wave-length of the light. The maximum intensity on the ordinary plates is 
about X 450, while the isochromatic plates have another maximum about X 560, and 
it is at these wave-lengths that the continuous spectrum extends furthest from the 
photosphere. The greatest extension is in Plates Nos. 13, 15, and 19, where it 
amounts to two-thirds of the sun’s apparent diameter, corresponding to a height 
above the photosphere of nearly 600,000 miles. 
