218 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1955 



actually depended upon the particular value of the zero point and the 

 resulting scale of distances. The unexpectedly large correction in the 

 scale is a shock to all theories involving the so-called cosmological con- 

 stant. We need not express regret that these theories were created — 

 they were fully justified by the esthetic value alone — but, from the 

 standpoint of economy of thought, the cosmological constant (equiva- 

 lent to a repulsion) must be suspended from active duty for the time 

 being and put in cold storage until new observational facts sound the 

 trumpet for its revival. It is rather doubtful whether this ever will 

 happen. 



The zero point of cepheid luminosities affects only the distances of 

 extragalactic nebulae. Within the galaxy, including the globular 

 clusters, a more reliable criterion of distance is offered by the known 

 luminosities of the short-period cepheids, the so-called cluster-type or 

 liR Lyrae variables. The average luminosity of these Population II 

 high-velocity objects does not depend so much upon space absorption, 

 and is well determined. They were too faint to be observed in the 

 nebulae by Hubble. In the Magellanic Clouds, whose estimated dis- 

 tances depended also upon long-period cepheids, persistent Harvard 

 Observatory searches failed to reveal cluster-type variables, a circum- 

 stance sometimes interpreted even as indicating the actual absence of 

 these objects. 



Now, as last, numerous cluster-type variables have been found in 

 the Magellanic Clouds (51), but about 1.3 mag (or 3.3 times) fainter 

 than expected from the magnitudes of the long-period cepheids. Thus, 

 the long-period cepheids are 1.3 mag brighter and all distances based 

 on them 1.8 times greater than was formerly assumed. The apparent 

 diameters and integrated luminosities of globular clusters in external 

 galaxies call for a similar correction (52), and independent support 

 for these conclusions is forthcoming from other sources (Baade). 



This, however, is not the whole story. The recession constant of the 

 nebulae depends entirely on the more distant objects, for obvious rea- 

 sons ; yet in these no variable stars could be observed. Their distances 

 were linked to the cepheid scale of the nearer galaxies through inter- 

 mediate criteria — the magnitudes of the brightest stars and of the 

 nebulae themselves. Both criteria are of a statistical nature and not 

 only involve various photometric errors, but also depend upon the true 

 dispersion (variety) of the magnitudes of the objects used as stand- 

 ards; the dispersions, and therefore the distances, seem to have been 

 underestimated by Hubble. A comprehensive survey of the problem 

 has been given by Behr (53). He concludes that those of Hubble's 

 intrinsic luminosities of the nebulae which are not based on variable 

 stars should be increased by 1.7 mag. Behr was not aware of the need 

 for adjustment of the cepheid scale of the nearest nebulae, and this 



