[ 24 ] 



is obtained by the fcrcw, which works vertically in the cock 

 C, Plate IV. Fig. I. and Plate II. Fig. IV. 



The juft diftance of the tool from the telefcope is Hiewn 

 by the diftind vifion of the images ; and in order to obtain the 

 fame diftance exadtly at all times, the fcrew D, in Fig. III. 

 Plate II. is moved, 'till its point juft prcffes againft the end of 

 the notched piece which refts on the bar, and is there fet by 

 the fcrew S (fame Jigure). Let now the fcrew W at the bottom 

 be turned, fo as to relieve the plumb-line, and let the tool be 

 removed ; then let the telefcope be inverted, and the tool 

 brought to the other fide, and placed upon the other bar and 

 cock, let the wire be made to fwing through and bifecl the 

 upper image; if now the telefcope has turned upon an axis 

 truly horizontal, the plumb-line will alfo bifed the lower 

 image ; if not, corred half the error as ufual, and repeat the 

 adjuftment 'till all error vanifhcs. 



Thus we fhall have the inftrument proved, without any addi- 

 tional weight on the axis; we have the advantage of a long 

 plumb-Hne, without parallax, or corpufcular attradion ; and all 

 error of the artift in the diameters of the pivots is rendered of 

 no confequence whatever. 



Apparatus 



