t)44 PHILOSOPHICAL TRANSACTIONS. [ANNO 1785. 



Atlas Coelestis, which must have the stars marked from the author's catalogue, 

 by a number easily added to every star with pen and ink, as I have done to mine. 

 The catalogue should also be numbered by an additional column, after that 

 which contains the magnitudes. I hope in some future editions of the Atlas to 

 see this method adopted in print, as the advantage of it is very considerable, 

 both in referring to the catalogue for the place of a star laid down in the Atlas, 

 and in finding a star in the latter whose place is given in the former. 



I would recommend a precaution to those who wish to examine the closest of 

 my double stars. It relates to the adjustment of the focus. Supposing the 

 telescope and the observer long enough out in the open air to have acquired a 

 settled temperature, and the night sufficiently clear for the purpose ; let the 

 focus of the instrument be re-adjusted with the utmost delicacy on a star known 

 to be single, of nearly the same altitude, magnitude, and colour, as the star 

 which is to be examined, or one star above and another below the same. 



The measures of the distances were all taken with a parallel silk-worm's-thread 

 micrometer, and a power of 227 only. They are not, as in the former cata- 

 logue, with the diameters included, but from the centre of one star to the 

 centre of the other. I have adopted these measures on finding that I could pro- 

 cure threads fine enough to subtend only an angle of about I" J 3'", and that by 

 this means there was no longer any great difficulty of judging when the stars 

 were centrally covered by the threads. However, I do not know whether these 

 measures, with stars at a considerable distance, may not be liable to an additional 

 error of perhaps l", owing to the remaining uncertainty in judging of their 

 exact central position while the measure is taking. 



The positions have all been measured, unless marked otherwise, with a power 

 of 460, adapted to an excellent micrometer, executed by Messrs. Nairne and 

 Blunt, according to the model given in the Philos. Trans, vol. 71, p. 500; but 

 with a great and necessary improvement of making the wheel d, d, of that figure 

 perform its whole revolution ; by which means the two silk-worms-threads may 

 be adjusted to a greater degree of exactness ; for if they are not placed so as per- 

 fectly to bisect the circle, the two threads will not coincide exactly after having 

 performed one semi-revolution, which they must be made to do with the utmost 

 rigour. I found the absolute necessity of this precaution when I came critically 

 to examine the positions of the Georgium Sidus, as they are given in table 3, 

 Phil. Trans, vol. 71, p. 497. The measures were affected with a small and 

 pretty regular error, which I was at a loss to account for ; and the distance of 

 this star being then totally unknown, I looked for the cause of the deviation at 

 first in a diurnal parallax of that heavenly body ; but soon found it owing to the 

 inconvenience before-mentioned, of not being able experimentally to adjust the 

 moveable thread to that critical nicety which I have now introduced and used in 

 all the angles of the present catalogue. 



