PROGRESS OF ASTRONOMY FOR 1891 AND 1892. 691 



Three lines have attracted more especial attention on acconnt of their 

 intimate connection with the susi)ected physical constitntion of tlie 

 star. '' These are (1) the bn<;h tureen line near h^, and the less refran- 

 .iiible edji'c of tlie hydrocarlxHi band; (2) the line near the chief 

 nebnlar line ATtOOd, and (.'>) tlie line near the pair of chromospheric 

 lines A4:92;3 and A4i)21. When the wave length of these lines, as ([noted 

 by the observers, are corrected for motion in the line of sight, and 

 arranged in a table, the mean valnes come ont very close to the wav^e 

 lengths of three notable pairs of solar chromosi)heric Hnes ; while mag- 

 nesinm and the hydrocarbons, as possible origins of line (1), are ex- 

 cluded by the absence from the lists of their inseparable companion 

 lines and Hutings. 



Line (2) is claimed by four observers for the chief nebular line, but 

 the weight of evidence seems to be against its nebular origin, and the 

 outburst would seem to be a vast chromospheric disturbance, a view 

 confirmed by Dr. Huggins' observation of the complete series of bright 

 hydrogen lines in the ultra-violet — the same that Hale and Deslaiulrcs 

 found in the solar chromosphere — but each with its dark companion. 



An interesting article advocating the meteoric theory in explanation 

 of the outburst is given by Prof. Lockyer in Volume 31 of the IS'ine- 

 teenth Century. The chromospheric theory of the near approach of 

 two stars is given by Dr. Huggins in the June numbei-, 1892, of the 

 Fortnightly Ifeview ; Seeliger's moditication of the meteoric theory is 

 translated in Astronomy and Astrophysics for December, and a single- 

 star chroinosjilicric theory is offered by Sidgreaves in the October 

 number of the Observatory. 



stp:llar spectra. 



Dr((i)er vataloyne of stelhtr spcctrn. — Volume 27 of the Harvard An- 

 nals contains a catalogue of the photographic spectra of more than ten 

 tliousand stars north of 25^ south declination. The i)hotographs were 

 taken with sm S-inch Voigtliinder lens, in front of which was placed a 

 ])rism S inches square, with a refracting angle of 13°. The edge of this 

 l>rism was so (ixed that the star's light was dispersed in declination, the 

 k'ngth of the spectrum being about a centimeter, aiul the star being- 

 allowed to trail slightly gave the spectrum a width of about a millimeter. 

 I^ach plate covered ](P s(pnire and the spectra of all stars to the sixth 

 magnitude were i)hotographed. The S])ectra are divided, for conveni- 

 ence, into a large number of classes — A BCD indicating varieties of the 

 first type; E to L, varieties of the second ty])e; M, the third type; N, 

 the fourth type; and O P Q S])ectra that do not resemble any of the 

 preceding types. One of the most important features of the work is 

 the method by which [)hotograj)]iic magnitudes have l)een assigned. 

 ''The (luantity measured in each case is the intensity of the spectrum 

 in the vicinity of the ( J line. Accordingly, when stars having ditferent 

 spectra are com])ared. the results will not be the same as if the entire 



