492 



Dr. and Mrs. Huggins. 



[May 19, 



These observers met also with double reversals, which gave a 

 triple character to the expanded single line. In one experiment, 

 when sodium carbonate was introduced into the arc the reversed 

 D lines were seen as a broad dark band with a bright diffuse band 

 in the middle. As the sodium evaporated the band narrowed, and 

 the bright line in the middle showed a second reversal within it. 

 This was a case of threefold reversal. 



There would seem to be little doubt but that the more or less divided 

 character — sometimes unsymmetrically — of the bright and dark lines 

 of the Nova, which appeared to be undergoing continual alterations, 

 was due to the incoming upon the broader lines of narrow reversed 

 lines, bright or dark, as the case might be. Provision must therefore 

 be made for conditions favourable for such reversals in any hypothesis 

 which is suggested to account for the phenomena of the new star. 



Waning of the Star. — The first record of this star was its appear- 

 ance as a star of the 5th magnitude on a plate taken at Cambridge, 

 U.S., on December 10, 1891. ~No star so bright as the 9th magnitude 

 was found at its place on a plate taken by Dr. Max Wolf on De- 

 cember 8. Combining the photographic magnitudes obtained at 

 Greenwich with the visual ones made at the University Observatory, 

 Oxford, and by Mr. Stone and Mr. Knott, we find that throughout 

 February and the first few days of March the light of the star de- 

 clined very slowly, but with continual and considerable fluctuations, 

 from about the 4'5th magnitude down to the 6th magnitude. After 

 March 7, the remarkable swayings to and fro of the intensity of the 

 light, set up probably by commotions attendant on the cause of its 

 outburst, calmed down, and the star fell rapidly and with regularity 

 to about the 11th magnitude by March 24, and then down to about 

 14*4th magnitude by April 1. On April 26, however, it was still 

 visible at Harvard Observatory, magnitude 14*5, on the scale of the 

 meridian photometer. 



We observed its spectrum for the last time on March 24, when it 

 had fallen to nearly the 11th magnitude. We were still able to 

 glimpse the chief features of its spectrum. The four bright lines 

 in the green were distinctly seen, and appeared to retain their rela- 

 tive brightness ; F the brightest, then the line near b, followed by the 

 lines about \ 5015 and X 4921. 



Traces of the continuous spectrum were still to be seen. Consider- 

 ing the much greater faintness of the continuous spectrum when the 

 star was bright on February 2 than the brilliant lines falling upon it, 

 we are not prepared to say that the falling off of the continuous spec- 

 trum was greater than might well be due to the waning of the star's 

 light. 



Professor Pickering informs me that on his plates " the principal 

 bright lines faded in the order K, H, a, F, h, and G, the latter line 



