394 



COMPARISON OF CATALOGUES OF STARS. 



The corrected residuals, in the case of Safford, in declination, of Boss and of New- 

 comb, are so nearly constant, for each group, when arranged in the order of declina- 

 tion, that it may be safely assumed that they are not a function of the right ascension. 

 In the case of Harvard College and of Safford, in right ascension, the evidence is 

 decisive that the residuals in certain groups vary with the right ascension. This is to 

 be expected in the case of the Harvard College observations, since the deviations of 

 the fundamental system upon which they depend, from the system of publication 

 XIX. are functions both of the declination and of the right ascension. The provisional 

 catalogue designated P is based upon Bessel's constants of precession. The same 

 remark applies to Safford's catalogue. For the conversion from the constants of 

 Bessel to those of Strove, we have the following relations between the annual varia- 

 tions in a and 8 for 1S7.">.0 : — 



Strove = Bessel +.001123 +.000129 tan 8 sin a. 



Struve = Bessel +.00194 cos a. 



It will be seen from these relations that the change in the right ascension in a 

 century may be as great as 0M1 for a star on the equator, while the change in decli- 

 nation will be insignificant, the maximum value being only 0".19. It is also apparent 

 that the residuals, especially in right ascension in the column Auwers minus P, are for 

 the most part due to the change from the constants of Bessel to those of Struve. 



For the more complete reduction to the system of Auwers. we have the following 

 additional data. 



Auwers minus Harvard College. 



