DEPARTMENT OF MERIDIAN ASTROMETRY. 213 



and computed from the changes of type. On the contrary, if the 

 general spectrum does not share in the changes of type shown by the 

 hydrogen spectrum, according to the results of Adams and Joy, then 

 the changes of color-index computed from the general spectrum are 

 practically zero, constituting an apparent outstanding discordance. 



A tabulation for a number of Cepheids also showed : (a) a progres- 

 sive trend of types with increasing length of period (known before), 

 the efifect being, however, very much more pronounced for short than 

 for longer periods, and (6) the apparently complete lack of a relation 

 between the range through which the type varies and the period. The 

 type range in the 22 Cepheids listed averages one type-interval (as F 

 to G), with only moderate deviations from the mean. 



Inasmuch as the surface conditions on Cepheids, i.e., in the com- 

 paratively shallow layer in which all of the light that reaches us 

 originates, are undergoing constant and apparently violent changes, 

 it seems probable that a considerable overlapping of the distinguishing 

 characteristics of types takes place. With good modern instrumental 

 equipment, applied especially to the brighter Cepheids, it should be 

 possible to determine approximately the extent to which Cepheid 

 spectra deviate from pure stellar types. 



REDUCTION OF OBSERVATIONS. 



The observations of the secondary observers made at San Luis 

 have been investigated for systematic errors, and their relation to the 

 results of the fundamental observers ascertained. For the night 

 observations of the five observers the latitude (omitting —30° 17') 

 and the refraction coefficients are respectively as follows: Roy 45 ''54, 

 0.9926; Sanford 45^48, 0.9950; Tucker 45^78, 0.9916; Varnum 45r51, 

 0.9942; and Zimmer45T49, 0.9936. It may be remarked, as has been 

 suggested elsewhere, that the discrepancy in the mean latitude may be 

 entirely due to personalities in determining the nadir ; or the zenith cor- 

 rection should be applied otherwise than symmetrically north and south. 



The collation of results is well advanced and the computation of 

 precession for the catalogue has been started. 



The first steps have been taken in the reductions of all the Albany 

 observations. The observations made during the last year of observ- 

 ing have been brought to the same stage as the previous Albany series. 

 The corrections for south minus north and eye and ear minus chrono- 

 graph were computed and applied. The first reductions of double 

 transits of circumpolar stars were made and preliminary places de- 

 rived for the 19 primary azimuth stars. Using these places, azimuths 

 were computed and plotted and curves drawn to represent them for 

 those series of the Albany observations where there were 12 and 24 

 hour groups of clock stars. These are furnishing the material for the 

 first investigation of the systematic correction in right-ascension 

 depending upon right-ascension now in progress. 



