16 
ASTRONOMY: H. SHAPLEY 
and ^'s more luminous than the very brightest h, which is an exceptional 
star; and of the 221 brightest stars in these outer regions of the cluster 
there are only four in color class h and twenty in class a. 
Fainter than the fifteenth magnitude the bluer colors become very 
common and there is an indication that the reddest stars may be almost 
entirely absent between magnitudes 15.5 and 17.0. At any rate, between 
magnitudes 16.0 and 16.8 only four stars out of the hundred measured 
have color indices greater than half a magnitude, and none is found 
redder than the sun. If the result is confirmed by later study it will be 
very important. It would suggest that this magnitude interval repre- 
sents in the cluster the special domain in the luminosity sequence of the 
normal first type star, and that the dwarf stars of the redder spectral 
types (including bodies comparable to the sun) will be found among the 
still fainter members of the system. 
From the foregoing considerations we conclude that not only are giant 
second type stars present in large numbers in the globular clusters, but 
also on the average among the four or five brightest magnitudes the 
brighter the star the redder its color. This condition is the inverse of 
that for the dwarfs, where, without much doubt, the cooKng stars are 
growing redder, fainter, and smaller with age. In the conventional 
scheme of the evolutionary transition of spectral types the present 
result introduces serious difficulties; for Russell's hypothesis, however, 
the difficulties are not so severe, but still the immediate ancestry of these 
giant red stars must be taken into account. More detailed considerations 
of the observations and their interpretation will appear in a forthcoming 
Contribution from the Mount Wilson Solar Observatory. 
The following references may be cited: 
Kapteyn, Groningen Publications, No. 14, 1904; No. 23, 1909; Trans. Internal. Solar 
Union, 3, 225 (1911). 
Eddington, Stellar Movements and the Structure of the Universe, chap, iv (London, 1914). 
Russell, Popular Astronomy, 22, 291 (1914). 
Scares and Shapley, These Proceedings 1, 487 (1915). 
Shapley, Ihid. 1, 459 (1915); 2, 12 (1916). 
