444 



SCIENCE. 



[N. S. Vol. XXIII. No. 586. 



'Catalogue of Variable Stars.' Tables of 

 observed maxima and minima for class II., 

 or long period stars, are now being printed. 

 Proof sheets of such tables for stars be- 

 tween and 6 hours right ascension are 

 now submitted to the society. These tables 

 give in successive columns, calendar date 

 of the observed phase, Julian date, magni- 

 tude, name of observer, reference, epoch 

 and residual from a period computed by 

 a formula following the name of the star. 

 Residuals have been calculated without the 

 use of a second term in the formula: by 

 plotting residual curves the form of cor- 

 rection to the period may often be found. 

 Residual curves of T Andromedce, o 

 (omicron) Ceti and S TJrsce, Majoris were 

 shown to the society. The residual curve 

 of S JJrsce Majoris, drawn from observa- 

 tions extending from 1843 to 1905, shows 

 the need of a sine term in the formula. 

 The residual curve of T Andromedce is in- 

 teresting in connection with the discussion 

 concerning the period of this star soon 

 after its discovery. Observations of o Ceti 

 have extended over a longer time than 

 those of any other long period variable. 

 The residual curve, computed from obser- 

 vations since 1596, shows many irregular- 

 ities and explains the difficulty of deducing 

 a formula to represent the phases of this 

 star. 



It is hoped that these tables, by bringing 

 together so much material in a convenient 

 form, may prove of some use in investigat- 

 ing the variables of long period. 



Systematic Study of Faint Stars: B. C. 



Pickering. 



This paper described work that has been 

 in progress at Harvard College Observa- 

 tory, during the past five years, in study- 

 ing the faint stars. The sky was divided 

 into 48 exactly equal parts, as explained 

 in detail in Harvard Annals, XIV., 477, 

 and used in various publications of the ob- 



servatory during the last twenty years. 

 Portions of the sky, one degree square, in 

 the centers of these regions, have been 

 chosen for detailed study. The epoch em- 

 ployed is 1900. A sequence of fifteen stars 

 of magnitudes 7.5 to 12.5 has been selected 

 and measured with the twelve-inch merid- 

 ian photometer, in each of these regions. 

 Photographs have been taken with two 

 hours ' exposure, using the twenty-four-inch 

 Bruce telescope, showing stars of magni- 

 tude 16, and brighter, and these have been 

 enlarged to a scale of 20" = 1 mm. Photo- 

 graphs of the spectra with small dis- 

 persion, showing stars to magnitude 11, 

 have been taken, and charts in and out of 

 focus have been obtained. Photographs of 

 these regions have also been made at the 

 altitude of the pole, and for comparison the 

 pole itself was photographed with the same 

 exposure upon the same plate. The photo- 

 graphic magnitudes of all stars in the 

 sequences mentioned above can thus be 

 determined on the same scale. The meas- 

 urement of photometric magnitudes from 

 isoehromatic plates is now in progress. 



Nearly all photographs of the thirty-six 

 regions north of declination — 30°, and 

 many of the others, have been completed; 

 the entire work would probably have been 

 finished, had the appropriation of 1902 

 been continued. 



Distribution of the Stars: Geo. C. Com- 



STOCK. 



This paper deals with the extent of the 

 visible universe and with the arrangement 

 of the stars composing it. The data avail- 

 able and utilized for the discussion are of 

 two kinds, (a) the number of stars per 

 square degree of the sky as shown by the 

 enumerations and gauges of Pickering, 

 Argelander-Seeliger, Celoria and the Her- 

 schels, extending from the brightest stars 

 down to the faintest visible in the Herschel 

 telescope, about the fourteenth magnitude; 



