220 CARNEGIE INSTITUTION OF WASHINGTON. 



sphere gives the observed elliptical distribution. This explanation is 

 supported by the observed high concentration in certain clusters of 

 blue stars and variables along the major axis, in close analogy with 

 well-known phenomena of the galactic system. 



Other cluster investigations have concerned the variable stars in 

 Messier 3 and Messier 5. In the latter a search is being made into 

 the colors and color variations of a group of variables with extremely 

 short periods, discovered by Bailey at Harvard. Catalogues of mag- 

 nitudes and colors have been pubUshed for Messier 11 and for four 

 fields in the neighboring dense galactic clouds. Work on a dozen 

 other clusters in and out of the Milky Way verifies the absence of 

 effectual scattering of Ught in space, and permits a provisional estimate 

 of the depth of the Milky Way as not less than 25,000 light-years in 

 some directions. A study of the magnitudes of 900 stars in Messier 3 is 

 nearing completion. In the search for very faint globular systems, 

 the general character of a number of faint clusters has been examined. 

 With a small photographic lens of 4 inches focal length the measure- 

 ment of the integrated light of all the brighter globular clusters is 



under way. 



Variable Stars and Miscellaneous Photometry. 



Several miscellaneous investigations of variable stars by Mr. and 

 Mrs. Shapley have been completed during the year, among which are 

 the derivation of a color curve for XZ Cygni, the study of the orbits 

 of some ecUpsing binaries, and an inquiry into the magnitude and color 

 of RU Bootis, one of the faintest of known periodic variables. Results 

 for the last indicate a distance from the galactic plane several times 

 greater than any heretofore found. Prehminary determinations show 

 that the faintest variable stars in the Milky Way are red. In conjunc- 

 tion with Dr. van der Bilt, new light elements have been derived for 

 W Ursae Majoris, an ecUpsing star remarkable for its very high density 

 and short period. The analysis involved 67 minima, covering an 

 interval of 13 years, and brought to light a small perturbation of the 

 period. Measurement of the magnitude of the ninth satellite of Jupiter 

 indicates that its diameter can not exceed 18 miles, if its albedo is com- 

 parable with that of the asteroids. A similar investigation of a faint 

 asteroid, whose orbit has been computed by Mr. Nicholson and Mr. 

 Shapley, shows it to have a diameter of approximately 3 miles; it is 

 accordingly the smallest planetary object for which dimensions have 

 yet been estunated. For the reductions connected with these various 

 studies, Mr. Shapley has had the assistance of Miss Davis and, for a 

 part of the time, that of Miss Richmond. 



STELLAR SPECTROSCOPY. 



The further development by Mr. Adams of the method of determin- 

 ing the absolute magnitudes of stars and their distances thi'ough a study 

 of their spectra has led to some modifications of the observing program 



