110 An Account of ike Repealing Circle, 



have used it for taking angular distances, as it has been in 

 astronomy ; and I have in this place but little to add to my 

 former remarks. The observer is not here, as in astronomy, 

 compelled to perform his operations in a short time : he may 

 take as much as he pleases, and in truth he will want not a 

 litde. Even by the improved method, it requires much time 

 to place the circle in the plane of the two objects: then fol- 

 lows the process of repetition ; then that of observing the ele- 

 vation or depression of the objects ; and lastly, measuring, if 

 they are near, or estimating, if remote, their distance from the 

 station, to enable him to correct the angle for the effect of 

 eccentricity of the telescopes. All diis must be done before 

 he has data for computing the horizontal angle ; and all this 

 is requisite for every angle that he takes ; even from the same 

 station. A gives the horizontal angle at once ; its parts are 

 perfectly concentric, and therefore the distance of the objects 

 is not concerned : the motion of the telescope, like that of a 

 transit, is truly perpendicular, therefore the elevation or de- 

 pression of the' objects comes not into the account. When the 

 adjusted instrument is placed with the axis of azimuth vertical, 

 the observer may proceed through all the angles of his station 

 as fast as he can observe and read them off. They will, how- 

 ever, be affected by the errors of graduation ; but it has been 

 shown how these may be diminished ; namely, where there are 

 three readmgs, by reversion ; and where there are only two, 

 in cases where extreme accuracy is required, by turning again 

 and again different parts of the circle towards the same object. 

 And although a fresh adjustment of the vertical axis thereby 

 becomes necessary at every step, this is an expedient attended 

 with far less trouble and loss of time than can be brought 

 about by repetition and its requisite accompaniments. 



Were it urged in favour of R, that by re))etition, the errors 

 of division being almost annihilated, it becomes more fit for 

 ascertaining the position of a star in the heavens ; because, in 

 observing one with A, the readings always take place on the 

 same divisions, and consequently the errors of those divisions 

 are charged upon die place of the star : this I should readily 

 gi-ant, were the two instruments equal in other respects. But 

 I have to observe that, in the present state of practical astro- 

 nomy, neither one instrument nor the other is at all (it for 

 such a purpose. It need hardly be remarked, that the resist- 

 ance and dragging of the centre- work, which I explahicd when 

 I endeavoured to trace to its sources the cause of fixed error, 

 will produce the same injurious effect on terrestrial measure- 

 ment as they do in astronomical observations. 



As a transit instrument I believe II was never thought of: 



indeed, 



