1869.] of August 18, 1868. 63 



till the commencement of totality, when they disappeared imme- 

 diately, leaving only a dull spectrum, still always perceptible and 

 continuous. A reversal of the spectrum, that is to say, an appear- 

 ance of luminous lines in the places occupied by the black lines did 

 not take place. When towards the end of totality, light clouds 

 covered the corona in such a manner as only to allow the light from 

 the red border and that fi'om the protuberances to pass, the spectrum 

 underwent a remarkable change, all the colours of greater refrangi- 

 bihty than the green disappeared, and the red extremity only 

 remained under the form of a spectrum interrupted by wide bands. 

 The first ray of the sun after totality instantly occasioned the 

 reappearance of Fraiinhofer's lines. 



Amongst the Enghsh observers Lieut. Herschel took the most 

 accurate observations. In a letter to Mr. Huggins, dated from 

 Belgaum, he gives the following vivid account of his discovery of 

 the constitution of the red flames : — 



"About a quarter of a minute before totality a thick cloud 

 obscured the sun. I had placed the slit of the spectroscope so as 

 to cross the crescent at about the vanishing point of the limb, and 

 was watching the narrow solar spectrum grow rapidly narrower. 

 You may conceive the state of nervous tension at this moment. 

 Whatever the corona was competent to show must in a few seconds 

 have been revealed — unless indeed it should so happen that a 

 prominence should be situated at that precise spot, in which case 

 the double spectrum would be presented. But the solar spectrum 

 faded out while it had still appreciable width, and I knew a cloud 

 was the cause. I went to the finder, removed the dark glass, and 

 waited — in that fever of philosophical impatience which recognizes 

 the futility of irritation, even wliile it chafes under the knowledge 

 of fleeting seconds — how long I cannot say, perhaps half-a-minute. 

 I can well recall the kind of frenzied temptation to turn screws and 

 look somewhere else, checked by the calm ticking of the clock 

 teUing of a firm hold of the right place, cloud or no cloud. Soon 

 the cloud hurried over, following the moon's direction, and therefore 

 revealing first the upper limb, with its radiating and, as I fancied, 

 scintillating corona, and then the lower limb. Instantly I marked 

 a prominence near the needle point in the finder. A rapid turn of 

 the tangent screw covered it with the point of the needle. Those 

 few seconds of unveiling were practically all that I saw of the 

 eclipse as a spectator. The instant the prominence was under 

 the needle point I returned to the spectroscope, A single glance 

 solved the problem in great measure. Three vivid lines — red, 

 orange, blue ! No others, no trace of a continuous spectrum. I 

 think I was a little excited about this time, for I shouted quite 

 unnecessarily to my recorder, ' Red, green, yellow,' quite conscious of 

 the fact that I meant orange and blue. I lost no time in applying 



