ASTRONOMY: H. SHAPLEY 
15 
1000 units of stellar distance, and it may be nearly ten times this amount. 
If we were to adopt the value of the absorption constant derived by 
Kapteyn, then the smallest color indices in the cluster should be in ex- 
cess of 2.5 magnitudes, and that for a star of spectrum M should be nearly 
five magnitudes. Actually not a single color index greater than two mag- 
nitudes has been observed. If we were to take Van Rhijn's value, the 
color indices in the cluster would all need to exceed a magnitude. 
It seems to be necessary to conclude that the selective extinction of 
light in space is entirely inappreciable, at least in the direction of the 
Hercules cluster. If we grant, on the basis of our data, a color excess of 
a tenth of a magnitude, and attribute it all to space absorption, the value 
of the coefficient can not then exceed -f- 0.0001 mag., an amount com- 
pletely negligible in dealing with the ordinary isolated stars. In the 
light of this result we are probably justified in assuming that the non- 
selective absorption in space (obstruction) is also negligible. 
The following references may be cited: 
Newcomb, The Stars, chap, xiv, New York, 1902. 
Turner, London, Man. Not. R. Astr. Soc, 69, 61 (1908). 
Kapteyn, Astrophys. J., 29, 46 (1909); 30, 284, 316 (1909); 40, 187 (1914). 
King, Cambridge, Mass., Ann. Obs. Harvard Coll., 59, 182 (1912); 76, 1 (1914). 
Jones, Observatory, London, 37, 402 (1914); Mon. Not. R. Astr. Soc, 75, 4 (1914). 
Van Rhijn, Dissertation, Groningen, 1915. 
Scares, These PROCEEb'iNGS, 1, 481 (1915). 
STUDIES OF MAGNITUDES IN STAR CLUSTERS. II. ON THE 
SEQUENCE OF SPECTRAL TYPES IN STELLAR EVOLUTION 
By Harlow Shapley 
MOUNT WILSON SOLAR OBSERVATORY. CARNEGIE INSTITUTION OF WASHINGTON 
Received by the Academy, December 9, 1915 
In many investigations of globular clusters the apparent magnitudes 
of the individual stars, which are the quantities observed directly, may 
be used without error as representing the absolute magnitudes. We 
may consider the stars in such a system to be at the same distance from 
the earth, and knowledge of that distance is unnecessary in discussing 
the relations between changes in the absolute luminosities of the stars 
and their colors or spectra. 
If, as is most commonly believed, the evolution of the stars progresses, 
chiefly through the agency of condensation and cooling, from the bluer 
spectral types to the redder, then, since it is definitely known that the 
intrinsic surface brightness of the blue stars greatly exceeds that of the 
red, we should normally expect to find that all the bright stars in the 
