-04 SIR a. BADEN-POWBLL ON THE TOTAL ECLIPSE OF THE SUN, 1896. 
defined than any other coronal feature. A large prominence (about latitude 45'^) is 
at the base of the group ; there is a decided thinning of the corona around it, and the 
lower rays, to a height of about 3', appear to bend over the prominence. 
A broad V-shaped opening, partly filled up with rays, separates the N.E. from the 
E. equatorial group. This is very large, reaching a greater height than any other 
portion of the corona (except the N.W. ray), and extending along the limb to about 
latitude 45° S., after which the group breaks up into a succession of broad ravs 
curving away from the S. pole, and forming the eastern portion of the S. polar rift. 
The detail on the eastern side of the corona is extremely complex and interesting, but 
a few jroints only can be noted. A small hook-shaped ray, about 2^' high, springs 
from a small prominence (about 18° N. latitude), and is distinctly bounded by a 
navrow dark space or outline. A large double-headed prominence (latitude about 
5° N.) is similarly outlined—the outline exactly following its contours. Apparently 
standing upon this prominence is a singular, dark, elliptical ring, about 2|-' by 
2', its longer axis nearly radially directed. From the top of the ring rises a thin, 
tapering ray, curved towards the south. South of the bright prominence are small 
rays which appear cut across by dark veins at heights of 2' and 3' from the 
limb. Further south the great mass is broken up in a manner entirely unusual. 
The solar corona usually appears composed of overlapping rays emanating from the 
sun, but here it appears to be also broken up by dark channels into flocculent-looking 
masses, giving to it somewhat of the curdled appearance of some parts of the nebula 
in Orion. The great mass is roughly divided into a northern and southern portion 
by an irregular ga]i or dark stream, commencing at the top of a mass of rays about 
6' from the limb; this gap turns towards the north, then curves east, and is 
lost at a height of about 17'. The base of the equatorial mass is filled by rays 
having much contorted forms. It is impossible to resist the impression that this 
portion of the corona is torn by violent storms or perturbations. 
Conclusions. 
1. The remarkable resemblance of the corona of 1896 to those of 1885 and 1886 
confirms the now recognized theory of periodic changes in the corona in accordance 
with variations in the solar activity, as shown by sun spots. 
2. The corona, of 1896, as will be seen by the foregoing description, shows decided 
evidence of a connection between corona and prominences. This was indicated in 
1893, but is still more striking in 1896. 
3. The corona of 1896 shows the hitherto unperceived features of dark streams or 
veins, which it seems impossible to regard as merely spaces between bright rays. 
The only features I have hitherto seen, which to any degree resemble them, are the 
comet-like markings on the E. side of the corona of 1871. The dark boidermg or out¬ 
line to some of the prominences I have not observed on any other eclipse photographs. 
