370 



errors are very different in these cases ; the first being at this 

 Observatory to the second as 4 to 5 ; and I even think the 

 stars ought not to differ in magnitude or colour : but if these 

 precautions be attended to, I am satisfied that very exact con- 

 clusions may be attained. Perhaps no stronger evidence of 

 this can be given than the result of a second examination, in- 

 stituted several years after certain alterations had been made 

 in consequence of the first one. 



" As the value of y was not very much less than the whole 

 e that I had obtained with the former eye-piece, notwithstand- 

 ing its low power, a new one was obtained from the late Mr. 

 Dollond, of 240, and very good. Nine lines were inserted by 

 Mr. Grubb, finer and more uniform than the previous seven ; 

 the clock, whose arc of repose had been four times that of 

 escape, had its weight changed from 4 lbs. to 1*75 ; and a mul- 

 titude of trees and shrubs were removed from the south of the 

 transit-room, as their evaporation was a manifest cause of un- 

 steadiness. All these, except the alteration of the clock, 

 tended to lessen the error ; that improved the rate, but it 

 made the beat less audible in high winds, and therefore would 

 increase u. 



" In this instance stars all of the first magnitude were se- 

 lected ; one of them, Fomalhout, I had formerly been unable 

 to use on account of its excessive fluctuations. They were 



a Lyra, g = + 38°. 39' . . . No. wires, 279 

 aAquike, + 8°. 29' . . . „ „ 251 



Fomalhout, - 30°. 27' ... • „ „ 278 



Giving 



(0 B -1022) 2 '= u 2 + y 2 x 1-63955 + z 2 x 1-76909 

 (0 8 -0925) 2 = u 2 + y 2 x 1-02225 + z 2 x 2-10826 

 (0 S -1162) 2 = u 2 + y 2 x 1-34559 + z 2 x 163-81153 

 Whence 



M = + 0"-0732; y = ±0 s -0554; z - + 8 -0049. 



