* 



SPECTROSCOPIC OBSERVATIONS. 83 



threads of continuous spectrum snapped and disappeared, and 

 totality had begun ; and I gave the signal " Go," and at the 

 instant the bright points disappeared also. 



I had experienced a great difficulty in the observation of the 

 last minutes of the partial phase. The sun was intensely bright, 

 and both dark glasses were none too much for the sun itself, 

 and one dark glass was rather too little for the spectrum. But 

 it became quite clear to me that it was risky to leave changing 

 all the dark glasses until the very moment of totality, as the 

 changes then would obviously be very rapid. I therefore dis- 

 carded one of the dark glasses from the direct image quite two 

 minutes before the total phase began, and one minute later 

 dropped the dark glass from the spectrum. In the next few 

 seconds I discarded the last dark glass from the direct image, 

 but kept that eye, the left eye, shut until after totality had 

 begun, observing only on the spectrum with the right. 



The spectrum even in the last minute was painfully bright 

 (I had tried again and again earlier to see if it were possible to 

 observe without the one dark glass on the spectrum, or without 

 the two glasses on the actual sun, but had found it impossible). 

 When therefore totality began I fancy my eyes were anything 

 but sensitive. But for a long time I could detect no trace of 

 the 1474 K " line. 



The spectrum of the corona was in the main a bright con- 

 tinuous spectrum of irregular brilliancy. The C, D 3 , F, and G 

 lines were seen as brilliant sharply denned semicircles, and 

 several very much shorter arcs were seen for a little time, the 

 6 triplet being one. But failing to detect " 1474 K," I turned 

 my attention after a time to the corona itself. 



When I turned back to the spectrum, the coronal line was 

 clearly seen, though very faint. Mid-totality was then past, and 

 the line was only seen on the west side of the moon. It was 

 traced through something less than half a circumference and 

 more than one-third say 140. 



It was much fainter than the arcs of the C, D 3 , F, and G lines, 

 but was broader, perhaps twice as broad. It was also diffused 

 on its outer edge, not sharp as the other four lines were. But 

 I could trace in it neither rifts nor rays. I estimated its 

 breadth (at the cry of " twenty ") as between i and of the 

 moon's diameter, or about 6' of arc. 



The eclipse now drew rapidly to its close ; at the cry of " ten " 

 the C, D 3 , F, and G lines were evidently brightening, and other 

 shorter arcs were appearing. In the next four seconds the 

 bright lines multiplied exceedingly, a number of bright threads 

 of continuous spectrum flashed out and ran together, a flood of 

 sunshine broke out and the eclipse was over. 



The time-keeper, Mr. Kamrao Subarao, clerk to Mr. Morris, 

 called out " Close cameras " at the fifth beat after the ten 



