326 PROCEEDINGS OF THE AMERICAN ACADEMY 



pains with the clock corrections ; so that now the framework of a good 

 evening's observation is the following : — 



( 2-3 standard polars. 

 1 2 " time stars. 

 Stars to be determined. 

 (" 2-3 standard polars. 

 ( 2 time stars. 



It sometimes happens, however, that a star to be determined is pre- 

 ceded by an observation upon five wires and followed by another upon 

 five wires of a close standard polar, like Polaris or X Ursae Minoris. 



In case the weather changes during the evening, I have considered 

 it safe to employ a one-sided determination of 7i and of the clock 

 correction whenever the mean of the transits of the two or three 

 standard polars is within an hour of that of the star to be deter- 

 mined. 



The standard polars used by preference are those of the Jahrbuch 

 within 10° of the pole; those between 5° and 10° receive three-fourths 

 weicfht. When it is necessary to go beyond 10°, stars between 10° 

 and 20° receive one-half weight. Such stars of the Jahrbuch as are 

 given in the following catalogue are beyond 10°, and the observations 

 here given have in no case been employed to determine the polar 

 deviation. 



The collimation has been determined by combination of level and 

 nadir, or, latterly, by two collimators ; one of these has an aperture 

 of 2h inches, and is an ordinary telescope with tube, pivots, level, and 

 micrometer ; the other is composed of a long-fociis object-glass of 

 five inches aperture and a temporary meridian mark, to be later re- 

 placed by a permanent apparatus, provided by the makers, but not yet 

 mounted. 



In the use of the catalogue of the Jahrbuch the only point to be 

 especially noted is, that polar stars whose places are relatively uncer- 

 tain have been, so far as possible, avoided. Their uncertainty has been 

 pointed out by Professor Auwers. the author of the catalogue ; and is 

 mainly owing to proper motion. The best stars are those which were 

 observed by Bradley twice or more. One pair of stars (7G Draconis 

 above pole, and Draconis 1 H. below, or vice versa) gave values of n 

 differing systematically from each other by over OM ; but it so hap- 

 pened that the mean correction, derived from later observations of 

 both, combined with a new investigation of their proper motions, was 

 almost exactly zero; so that I made it a rule to observe the pair 



