514 



CIRCLE. 



With an instrument of such moderate dimension*, it 

 may reasonably be expected, that different itari will give 

 different result*, therefore at many of these standard (tan 

 a* pOMibk ihould be observed every night, and the mean 

 difference of the obiervcd and computed place should be 

 taken as the error of collinmion, to be applied to the ob- 

 servations of the other start, whose north polar distances 

 it is the object of the observer to obtain. It in almost 

 unnccettary to add, that if the sun, moon, or planets, 

 make part of series of obsenrations, the same error of 

 collimation is to be applied to them. 



Having finished one aeries of results in the above po- 

 ition of the telescope, or rather level, it should be un- 

 clamped and changed to a new position, for instance, so 

 to mark 10* when the telescope points to the pole. 

 In every observation now, the instrument will mark 10 

 more than the polar distance of the star, besides a new 

 error of collimation, which is to be investigated by a 

 comparison with the Greenwich Catalogue, the same as 

 before. 



By comparing these results with those obtained in the 



first position, the observer will be able to form some opi- 

 nion of the degree of confidence to which his observation] 

 are entitled. He may now change the position of liij in- 

 strument another 10, and so proceed till he is satisfied of 

 the accuracy of his catalogue. We are fully convinced, 

 that, with the exception of Polaris and ft Ursi Minoris, 

 a much more accurate catalogue may be made by this 

 method than by the operation of repeating, which in fact 

 does not appear to have ever been very successfully em- 

 ployed for stars south of the zenith, even by the French 

 astronomers, notwithstanding their partiality to this in- 

 strument. The French instruments being very indiffe- 

 rently divided, the above method of observation would 

 not suggest itself, because the error of a single observa- 

 tion would probably be very considerable; but in instru- 

 ments of Mr Troughton's construction, it would rarely 

 occur that the error of division would amount to 5", 

 the error of observation, bisection, level, &c. &c. must 

 be common to every method that can be devised. 



We shall now conclude this article with Mr Pond's Ca- 

 talogue, referred to in the preceding page. 



Circle. 



Mr Ponft Catalogue of the North Polar Distances of 44 of the principal Fixed Stars, deduced from I , 



Observations. 



