284 SCIENTIFIC RECORD FOR 18S2. 



at least one magnitude, and be concludes that we should look with doubt 

 upon red variable stars where the variability is less than a magnitude 

 or so. 



Parallax of the star Bradley 3077. — Dr. Gylden has lately published 

 the results of his micrometric comparisons of this star with neighboring 

 ones. His memoir gives the parallax as 0".28 ± 0".045. Professor Bru- 

 enuow found by a similar method the parallax to be 0".07 i 0".02. 



In this connection a recent i)aper by Dr. Backlund, of Pulkova, is in- 

 teresting, as it gives the reduction of some observations begun by Dr. 

 Wagner on the same star. These observations were made by means of 

 differences of R. A. with the transit instrument, and they were discon- 

 tinued as the comparisou stars were too faint to observe during the 

 whole year. From the incomplete series Dr. Backlund has obtained the 

 following results: 



Twenty-seven comparisons with star a give + 0".20 ± 0".080. 

 Thirty-two comparisons with start give -f 0".21 ± 0".078. This star 

 has thus a small parallax. 



Parallaxes of Alpha Lyroe and of Gl Cygni. — During 1800 and 1881 

 ProlJessor Hall, of the Naval Observatory, observed for the parallaxes 

 of these two stars on 77 and 66 nights, respectivelj^, using the 26-inch 

 equatorial and micrometer I, with an eyepiece magnifying 383 times. 

 The results have been printed in the Washington observations for 1879. 

 The observations are of extreme precision, the probable errors of a single 

 distance in the case of A?p/mi/2/rte being 0".075 and in that of 61 Cygni 

 0".081. 



A temperature coefficient was found for the screw from observations 

 of Ad made during the series. The range of temperature was 42°r. If 

 this coefficient be used to reduce the 900 observations of transits made 

 in 1875, in which the range of temperature was 47° (Wash. Ast. Obs. 

 1877, App. I), the probable error of the resulting Eo is ± 0".0020. If the 

 temperature coefficient is omitted, the probable error is less than this. 

 As far as this goes, it would imiicate that the coefficient is doubtful. 



At any rate, its introduction will not affect the values of the resulting 

 parallaxes, which are : 



Alpha Lyrce 0'M80 i 0".006 



61 Cygni 0".478 i 0".014 



It will be interesting to compare these results with preceding ones. 

 For 61 Cygni we have— 



Bessel . 0". 357 



Bessel & Schliiter .536 



Johnson, 1st .526 



Johnson, 2d . 192 



Struve .511 



Auwers . 564 



Peters . 349 



Ball .465 



