230 CARNEGIE INSTITUTION OF WASHINGTON. 



characteristics were found to have bright Hnes (p. 274). The spectra 

 of the companions of Capella, Antares, and Sirius have been photo- 

 graphed, and much miscellaneous spectroscopic work has been done 

 (p. 274). 



Mr. Merrill has now determined 101 radial velocities of long-period 

 variables, 91 of which resemble o Ceti, while 10 have spectra of a new 

 type. 83 of these stars, of the Md (o Ceti) type, give from their 

 bright lines the following values for the solar motion: V = 56 km., 

 A =274°, D=+44°. The average residual velocity of 31 km. is the 

 greatest known for stars of any spectral type (p. 269). Several addi- 

 tional nebular lines have been found in R Aquarii; the displacements 

 of the nebular and stellar lines are being studied. The prominent 

 bright lines of the Md stars have been observed after maximum in 

 the spectrum of R Cygni, a variable with very different absorption 

 spectrum. Humason and Merrill have discovered during the year 

 more than 50 stars with bright Ha line (p. 275). 



The radial velocities of 28 R-type stars have been found by Mr. 

 Sanford to range from +60 to —400 km. The algebraic mean freed 

 from solar motion is —17 km. (p. 269). 



Mr. Hubble has photographed the spectra of about 150 nebulous 

 stars between 10.5 and 14.0 magnitude. Bright-line stars in extended 

 nebulae are nearly all of types Oe5 and BO, while those giving an absorp- 

 tion spectrum are Bl or later, averaging about B4. The central stars 

 in six large planetary nebulae have spectra intermediate between the 

 Wolf-Rayet and Oe5 types. Combining these results with those of 

 Wright on smaller planetaries, we have the following sequence of types 

 for stars involved in galactic nebulae: small planetaries, probably 

 Wolf-Rayet ; large planetaries, between Wolf-Rayet and Oe5 ; extended 

 bright-line nebulae, Oe5 to BO; extended nebulae with absorption 

 spectrum, Bl to A3. This suggests the possibility that only the earli- 

 est and hottest stars may be capable of exciting bright-line emission in 

 surrounding nebulosity. Near a critical tj-pe, about Bl, the bright 

 lines fade rapidly and an absorption spectrum soon predominates, 

 perhaps representing the reflected light of the star, as indicated by Sli- 

 pher. This tentative suggestion is being thoroughly tested (p. 254). 



The study of spectra of Novae, old and recent, has been continued by 

 Messrs. Adams and Joy. Except for decreased brightness of the ring 

 relatively to the central star, the extraordinary spectrum of Nova 

 Aquilae 1918 has changed but little during the year. Nova Cygni 

 1920 showed important changes within two weeks, both in type and 

 velocity. T Coronae, the Nova of 1866, was found to have an absorp- 

 tion spectrum closely resembling that of a typical giant Ma star, 

 with bright hydrogen and X4686 lines (p. 274). 



Kapteyn has developed during the year a theory of the arrangement 

 and motion of the entire stellar system. From his results on the dis- 



