428 , PHILOSOPHICAL TRANSACTIONS. [anNO 1747-8. 



it to be nearest to 40'; which number I therefore make use of in the following 

 computations. 



The divisions or points on the limb of my sector are placed 5 minutes of a 

 degree from each other; and are numbered so as to show the polar distances 

 nearly; the true polar distance exceeding that which is shown by the instrument, 

 about l' 35'^ When I first began to observe, I generally made use of that 

 point on the limb which was nearest to the star's polar distance, without regard- 

 ing whether it was more northerly, or more southerly than the star: but as it 

 sometimes happened that the original point, with which I at first compared the star, 

 became in process of time pretty remote from it, I afterwards brought the plum- 

 met to another point, that was nearer to it ; and carefully examined what number 

 of revolutions of the screw of the micrometer, &c. corresponded to the distance 

 between the different points that I had made use of: by which means I was able 

 to reduce all the observations of the same star to the same point, without sup- 

 posing the several divisions to be accurately 5' asunder. 



I have expressed the distance of each star from the point of the arc, with 

 which it was compared, in seconds of a degree and 10th parts of a second, 

 exactly as it was collected from the observations ; though I am sensible that the 

 observations themselves are liable to an error of more than a whole second; 

 because I meet with some that have been made within 2 or 3 days of each 

 other, that differ 1", even when they are not marked as defective in any 

 respect. 



It would be too tedious, to set down the whole number of the observations 

 that I have mide; and therefore I shall give only enough of them to show their 

 correspondency with the beforementioned hypothesis in the several years in which 

 any were made of the stars here recited. When several observations have been 

 taken of the same star, within a few days of each other, I have either set down 

 the mean result, or that observation which best agreed with it. I have likewise 

 commonly chosen those that were made near the same season of the year, in 

 such stars as gave me the opportunity of making that choice; particularly in y 

 Draconis, which was generally observed about the end of August or the begin- 

 ning of September ; that being the usual time when I went to Wansted, on pur- 

 pose to observe both that, and also some of the stars in the Great Bear. But the 

 weather proving cloudy at that season in the year 1744, prevented my making 

 a single observation, either of y Draconis, or any other star, while I was there; 

 which is the cause of one vacancy in a series of 20 succeeding years, in which 

 that particular star had been observed. Such stars as were either not visible in 

 the day-time, towards the beginning of September, or came at such hours of the 

 night as would have incommoded the family of the house in which the instru- 

 ment is fixed, were but seldom observed after I went to reside at Oxford : which 



