MOUNT WILSON OBSERVATORY. 243 



red stars of the same type (with the possible exception of the M's) as 

 Adams, Kohlschiitter, Seares, and others have shown, the giants are 

 much redder than the dwarfs. To test cluster-stars for luminosity, 

 well-known giants and dwarfs of the solar neighborhood have been pho- 

 tographed on the same plate wdth the cluster. A direct comparison of 

 the chromatic images of giants, dwarfs, and cluster-stars of the same 

 spectral types indicates that the cluster-stars are giants. The work is 

 being continued and an attempt mil be made to place the method on a 

 quantitative basis, though the difficulties pointed out by Mr. Seares in 

 his discussion of the relation of color to intrinsic luminosity remain to 

 be overcome. 



The detailed photometric catalogue of 848 bright stars in M3 has 

 been published by Mr. Shapley and Miss Davis. It affords valuable 

 material for the discussion of the properties of stars in globular clusters, 

 and serves to corroborate rather than to modify the principal results 

 obtained from the earlier and similar photometric study of M 13. 

 Omitting variables, the spectral curve of M 3 shows a maximum fre- 

 quency near color-class / 5 ; the magnitude curve shows a maximum 

 frequency for photovisual magnitude 15.5 (near absolute magnitude 

 zero) and, as usual, is far from being symmetrical. The relation of 

 color-index to magnitude shows the customary decrease of index with 

 decreasing brightness. 



A more comprehensive discussion of stars of the highest luminosity 

 has been based upon the analysis of 1,152 giant stars in 9 clusters, for 

 which distances, colors, and apparent magnitudes have been obtained 

 at Mount Wilson during the last five years. From this work it appears 

 that very few stars in clusters are brighter photovisually than -4, and 

 none has been found to exceed photovisual magnitude —6 absolute. 



The problem of dwarf stars in clusters is one that merits attention 

 because of its bearing on the relation of globular to open clusters and of 

 open clusters to the star-fields of the galactic system. Using an inten- 

 sifying device in connection with the 100-inch reflector, thus gaining 

 about one magnitude in the photographic limit of the telescope, Mr. 

 Shapley believes he has recorded the faintest stars existing in the globu- 

 lar cluster M 22, and there are indications that the lower limit of bright- 

 ness has also been approached in the Hercules cluster, M 13. In both 

 of these systems stars of solar brightness appear to be relatively infre- 

 quent, but it is of course impossible to say that much fainter dwarfs 

 are entirely absent. 



Computations based upon Holetschek's visual magnitudes and upon 

 determinations of distance by Mr. Shapley have yielded the total 

 intrinsic luminosities of 40 globular clusters. The average absolute 

 magnitude is —8. 8 ±0.5, corresponding to 275,000 times the Ught 

 emission of the sun. An independent method of measuring relative 

 distances of clusters is afforded by the small dispersion in the absolute 



