1904.] on Spectroscopic Studies of Astrophijsical Problems. 487 



solar atmosphere, and in drifting away has fallen down like the lava 

 of a volcanic eruption down the slopes of the crater. High in the 

 solar atmosphere there is no absorbing vapour between it and our 

 spectrograph to dull its glow. Falling down, it loses its central part 

 by the absorption of attenuated vapour above it, and shows this 

 forked appearance of the extremity at a distance from the spot. 



The neighbouring hydrogen line, H^, is also clearly reversed, 

 and probably quite similarly, but it is much weaker and shows only 

 the form of the brighter parts of the Ca line. It appeared on four 

 negatives, exposed within the same half hour, on September 9th, 

 1893, to the spot on the eastern limb. Possibly these are the only 

 photographs of a reversed He on the solar surface. 



I will now pass to the second instrument left to me by Fr. Perry, 

 the stellar spectrometer. It was designed for visual work, fitted 

 with every convenience for mapping faint spectra in the dark. I 

 spent hours with it before accepting the conclusion that I could have 

 no confidence in my direct eye-observations. Photography was a 

 necessity for me ; and photography was impossible without altera- 

 tions, which would have exposed me to the reproach of having ruined 

 an excellent visual spectroscope. I had to make use of tools of my 

 own, for I could not give up the stars. These had always been to 

 me my wonder-land ; and to gain ever so little knowledge of them- 

 selves and their movements, through the delicate touches of light, was 

 more attractive than any discoveries on the surface of a star so near 

 at home as the sun. I will not detain you with a description of my 

 tools further than to say that I have never been ashamed to show and 

 explain the inexpensive additions to an excellent 4-inch refractor 

 which have enabled a small visual j^rismatic camera to furnish the 

 photographic material for a spectroscopic study of the problematic 

 variable star of the Lyre. 



The material is not so abundant as could be desired. The 

 phenomenally cloudy state of our night skies during the past two 

 years is enough to account for the deficiency, without stopping to 

 enumerate and explain all the adversities which have contributed to 

 reduce the collection of photographic plates to the 54 which go to 

 form what has been called at the Observatory the spectrographic chart 

 of ^ Lyrae. 



The star is a well-known variable, and its light-changes have 

 attracted the attention of observers for over a century. Its charac- 

 teristic, as a variable, is an alternately greater and less loss of light 

 between successive equal maxima, the cycle of changes being com- 

 plete in nearly thirteen days. The periodic changes are represented 

 by the curve of one cycle projected on the screen, in which it will 

 be observed that the four extremes, two maxima and two minima, 

 divide the cycle into approximately equal parts. This might have 

 suggested a binary star of short period, in a circular orbit lying in 

 our line of sight. The two stars would then alternately eclipse one 

 another, and one being brighter than the other the consequent 



