184 



SCIENCE. 



[N. S. Vol. XXIV. No. 606. 



of only two minutes. The scale of the photo- 

 graphs, however, is so small that great magni- 

 fication is necessary, and many details must 

 be obscured, although the definition at the 

 center of the small plates used appears to be 

 good. There must be some practical limit to 

 the decrease in the ratio of focus to aperture, 

 and Schaeberle has apparently been working 

 near, if not beyond it. Vogel gives some won- 

 derfully good results obtained with a Schmidt 

 mirror, using an aperture of 24 cm. which 

 makes the ratio of aperture to focus 1:3.86. 

 With the full aperture of 41 cm. and an ex- 

 posure of ten minutes, many more stars were 

 photographed than are visible in the great 

 Lick refractor of more than 91 cm. aperture. 



SOME CONSIDERATIONS REGARDING THE NUMBER 

 OP THE STARS. 



The above title is the subject of a paper in 

 the May number of the Monthly Notices of 

 the Royal Astronomical Society, It is writ- 

 ten by Miss Winnifred Gibson, B.Sc, and 

 communicated by Professor Karl Pearson, 

 F.E.S, 



The problem is one of the deepest interest, 

 but its solution is rendered difficult, if not 

 impossible at the present time, by lack of 

 fundamental data. The distribution of the 

 stars would be readily determinable, if the 

 parallaxes of a sufficient number had been de- 

 termined. After confessing the lack of such 

 materials. Miss Gibson, nevertheless, proceeds 

 to an elaborate discussion of such data as 

 exist. She arrives at the conclusion, for the 

 brighter stars, that ' There is no sensible rela- 

 tion between magnitude and parallax.' The 

 chief trouble with this conclusion is that it 

 is derived from entirely inadequate data. The 

 stars, whose parallaxes are taken into consid- 

 eration, are of the ninth magnitude, and 

 brighter. Of such stars there are in the sky 

 about 150,000. Por obtaining any law of rela- 

 tion between the magnitudes and distances of 

 all these stars, the parallaxes of 72 stars, about 

 one in 2,000, are available. Neither is there 

 evidence that this small number is distributed 

 by chance among the stars, nor that the par- 

 allax in many cases is sufficiently exact for 

 purposes of discussion. Results based upon 



such meager data must have little or no value, 

 however skillful and elaborate the treatment 

 may be. Not much more can be said of that 

 part of the discussion relating to proper mo- 

 tions and colors. 



The latter part of the paper is concerned 

 with the relation of magnitudes and frequen- 

 cies, and gives results, which are in substantial 

 agreement with those of Professor E. C. Pick- 

 ering, recently published, although Miss Gib- 

 son prefers a somewhat different formula. 

 The author confines her attention to the 

 brighter stars. The problem becomes at once 

 more interesting and more difficult as it is 

 extended to the fainter stars. Miss Gibson 

 complicates the problem by assuming that 

 magnitude is largely a physiological phenom- 

 enon, and hence that * It is hardly conceivable 

 that a scale of pure magnitudes could have 

 any meaning in physical nature.' This is 

 quite erroneous, since the determination of 

 magnitudes, whether by visual or photographic 

 methods, is simply the determination of the 

 intensity of certain radiations, which surely 

 have an objective reality based on physical 

 conditions. 



Solon I. Bailey. 



Hasvaed College Observatoey. 



RECENT VERTEBRATE PALEONTOLOGY. 



From the Brazilian Coal Commission, 

 through Dr. I. C. White, state geologist of 

 West Virginia, the American Museum of 

 Natural History has just received the gift of 

 a number of natural casts of a small Permian 

 reptile closely related to the Mesosaurus of 

 South Africa. These Brazilian remains have 

 been carefully studied by J. H. McGregor 

 for a memoir which will be published by the 

 commission, and which for the first time gives 

 us a complete knowledge of the anatomy of 

 this extremely ancient reptile. It is related 

 to the genus Stereosternum Cope, also from 

 Brazil, which Baur made the type of the order 

 Proganosauria, under the impression that these 

 were very ancient and generalized reptiles. 

 It is true they are very ancient but are not 

 generalized, they are rather already consider- 

 ably specialized for aquatic life, as shown in 

 Dr. McGregor's restoration of the skeleton and 



