OF ARTS AND SCIENCES. 40 



o 



curves. It will not escape attention, that in this case, at least, the val- 

 ues in the two tables differ systematically at certain points. It is the 

 experience of the writer that the periodicity thus introduced some- 

 times amounts to one fourth or one fifth as much as the systematic 

 deviation of the catalogue compared from the normal system. 



In the paper on " A Comparison of the Harvard College Observa- 

 tory Catalogue of Stars for 1875.0 with the Fundamental Systems of 

 Auwers, Safford, Boss, and Newcomb," the method here described 

 has been essentially followed. But in order to reduce the magnitude 

 of the residuals with which we have to deal, the computed corrections 

 depending on both the right ascension and the declination were sub- 

 tracted from the original residuals, and the values of d/Ja and d/18 

 which remain were treated in the way above described, giving the 

 corrections found in Table III. Notwithstandincj the criticism of 

 Professor Safford that this method is unusual, I must maintain that 

 it gives nearer approximations to the true corrections than can be 

 obtained in any other way. 



Professor Safford has given an exhaustive discussion of the most 

 probable values of the right ascensions of a list of stars given in the 

 Memoir to which reference has been made. This discussion involves, 

 however, a knowledge of the systematic errors of the catalogues com- 

 pared. It will be interesting to compare with his results the final 

 results of the Harvard College observations of these stars, in which 

 there is a direct dependence upon the system of Publication XIV., 

 and in which, therefore, there is no need of applying corrections for a 

 systematic deviation from the assumed system. An experience of 

 thirteen years has shown that the Harvard College Meridian Circle 

 can be relied upon to follow very closely the fundamental system 

 chosen in a series of differential observations. 



Two stars of the list have erroneous values in the final catalosrue. 

 For /3Lyrte, the right ascension should read IS'* 45"' 27'.851, in- 

 stead of 27^774. 



In the case of v Pegasi, there is a misprint in the volume for 1872 

 of 23*' IS™ 59^374 for 59^.574. 



By a reference to the original manuscripts it is found that both of 

 these errors had been corrected, but by some mistake the corrections 

 were not made on the sheets prepared for the printer. There were 

 about a dozen errors of this kind, but all of them seem to have been 

 corrected in printing except these two. 



For the remaining stars the deviations from the positions given in 

 Publication XIV. are given below, except for the star 1 H. Draconis, 



