209 
the description I had seen of its “green velvety’? appearance. 
The green was like that of the olives that were commonly 
on the dinner-table in that neighbourhood, but greener and less 
brown. The clouds drifting across the moon were perceptible, 
and the mountains in the moon were not. 
5°. JI examined the corona for polarization near the ap- 
parent upper surface of the moon; I found that it was ap- 
parently polarized and that the diaphragm was vertical when 
the two images were of equal intensity. I did not determine 
its plane, but it must have been inclined at 45° to the horizon 
and therefore neither radial nor tangential. 
6°. Time was nearly up. I took my eye from the telescope 
and for a few seconds saw the general appearance of the corona 
and prominences. I distinctly noticed the quadrilateral shape 
ef the corona and that the sides of the quadrilateral were 
roughly horizontal, and vertical. The greatest extension was to 
the N.W. and was about 2 of the moon’s diameter. 
7. I returned to the telescope to watch for the final dis- 
appearance of the corona after totality. This occurred at 
12", 10”. 30° (12. 17. 50. G. M. T.). 
I had seen the corona for 2”. 50°, about ? of a minute longer 
than the totality. 
8°. Immediately after totality I hastily made a sketch of 
the impression scarcely gone from my eye of what I saw with- 
out the telescope ; this was very rudely and almost automatic- 
ally done, but agrees very closely with Mr Browne’s sketch. 
Mr Browne was standing by my side to make the sketch: 
he did not see mine nor I his till both were completed. 
Conclusions. 
I draw the following conclusions from the observations made 
during the eclipse and during the previous week. 
1°. All observations of polarization made with telescopes 
that have not been previously tested for blackness inside, are 
utterly untrustworthy. 
