250 CARNEGIE INSTITUTION OF WASHINGTON. 



color-curve, based on several hundred observations, will be completed 

 within a few months. The importance of the studj^ of such color- 

 changes lies in the fact that a conspicuous variation in the spectral 

 tj-^pe has been found to be the underlying cause, and an analysis of 

 these changes promises to throw further light on the nature of Cepheid 

 variation. Some new variables, probably of the Cepheid type, have 

 been found in the globular cluster Messier 9. 



The period of the well-known eclipsing star U Cephei has been 

 studied upon the basis of more than 20,000 photometric measures 

 by Wendell, Chandler, and others. An interval of 33 j^ears is covered 

 by the observations, and throughout nearly half of this time the 

 detennination of variations in the period has a probable error of less 

 than half a minute of tone. The high accuracy permits the conclusions 

 that a linear formula for the light elements is no longer applicable and 

 that comphcated variations in the elements exist. Orbits have been 

 computed for the eclipsing binaries Y Camelopardalis and TT Lyrae, 

 and revisions of several other orbits have been made. The determina- 

 tion of photographic and photovisual light curves for TW Andromedae 

 is nearl}?^ complete. 



STELLAR SPECTROSCOPY. 



The stellar spectroscopic work during the year has been continued 

 by Mr. Adams along two distinct lines. The first consists of radial- 

 velocity observations of lists of stars selected with certain definite 

 problems of stellar motion in mind. The second consists in the 

 development of the method of stellar classification and the derivation 

 of relationships between spectral characteristics and the intrinsic 

 brightness of stars. It is evident that a satisfactory method of deter- 

 mining absolute stellar magnitudes would, apart from other considera- 

 tions, be of great importance in its bearing on radial-velocity observa- 

 tions, since it would at once provide the means for discussing the 

 relationship between the motions and the probable masses of stars. 



A large part of the observational work has been based on two lists 

 of stars. The first consists of about 160 stars of visual magnitudes 7.5 

 to 9.0 with small proper motions; this is now about two-thirds com- 

 pleted. The second list is made up of the fainter stars with measured 

 parallaxes. For many of these a camera of 18.3 cm. focal length has 

 been employed, the radial velocity results obtained from these small- 

 scale photographs showing a very fair degree of accordance. Numer- 

 ous observations have been made on stars of known velocity, on bright 

 stars of large and small proper motion, and on miscellaneous stars, 

 including some of the fainter variables. 



The number of photographs obtained during the year is 650, distrib- 

 uted as follows: 



Small proper-motion stars 249 



Parallax and larKC-motion stars -74 



M iscellancovis 1 - ' 



