DEPARTMENT OF MERIDIAN ASTROMETRY. 193 



lost motion. As soon as the temperature rises or falls beyond a certain 

 point a new maximum or minimum is established. The observed data 

 seem to confirm the existence of such a term. 



The former practice was to group the colhmations over a period 

 within which the temperature coefficient seems to satisfy the collima- 

 tions. The advantages in using the lag term consist of the longer 

 range over which the colhmation can be applied by formula and the 

 saving of labor by applying a term which is constant for several weeks. 



THE VARIATIONS IN SPECTRAL TYPE OF THE FOURTH CLASS VARIABLE STAR 



1 CARIN/E. 



A paper was prepared and read at the twenty-first meeting of the 

 American Astronomical Society at Albany by Dr. Albrecht on the vari- 

 ations in spectral type of the fourth class variable star I CarinoB, of 

 which a brief abstract follows. 



When the discovery was made in 1906 that many of the lines in 

 stellar spectra show slightly different wave-lengths in the different 

 stellar types and that these differences progress with the regular type 

 sequence, it was pointed out that similar changes in wave-length 

 which would be progressive with the phase of light-variation might be 

 found in the spectra of individual variable stars. Such periodic changes 

 in wave-length, if found, would offer a quantitative method for the 

 determination of actual changes in spectral type synchronously with 

 the changes in Hght. The limited amount of material then available, 

 principally for the star ?; Aquilce, "showed very strong indications of 

 just such variations in the positions of certain fines." At Dr. Albrecht's 

 suggestion. Director Campbell, of the Lick Observatory, asked Dr. 

 Heber D. Curtis, then in charge of the D. 0. Mills Observatory in Chile, 

 to secure for this study series of spectrograms of the fourth-class vari- 

 able stars I Carince and k Pavonis. The present paper discusses the 

 measures of the spectrograms of I Carince which were secured at that 

 time. The methods employed are described in the Astrophysical 

 Journal, vol. 33, p. 130, 1911. The details of the investigation will be 

 pubfished later. Briefly summarizing, it may be stated that, as in the 

 case of T} Aquilce so also for I Carince, variations in spectral type are 

 shown to occur. For this star the range of variation is seven-tenths of 

 a type interval, the type being F 7.8 near fight maximum, G 2.9 at a 

 phase intermediate between maximum and minimum, and G 4.8 near 

 fight minimum. 



PERSONAL EQUATION OF SAN LUIS OBSERVERS. 



The tabulation of the residuals of the San Luis transits shows decided 

 systematic differences between the results obtained by the different 

 observers dependent upon decfination. Mr. Varnum has drawn up the 

 results and discussed them. The mean from the three fundamental 

 observers was taken as a standard and the error of each observer rela- 



