156 EASY AND ACCURATE METHOD of 



ed to a piece of board, and this board to a three or four 

 pound fhot, or fmall hand-granade, when either ofthefe^ 

 may be conveniently had. 



Next take a piece of triangular board of about four 

 inches on the llde, and through this cut a triangular 

 mortice of about two inches on the fide. Near the cor- 

 nel's of this board let there be inferted three fmall nails 

 or pieces of wire, to ferve as feet for it to ftand on. 



Method of making the Adjuftment, or finding the ^antity of the 

 Index-Error. 



At any time when the fun fhines, fet your triangular 

 board on a table, the cill of a window, or any other con- 

 venient ftand expofed to the fun, and place the ball with 

 the piece of glafs, on the triangular mortice j which, 

 touching the ball only in three points, will confequently 

 keep it fteady in any pofition. Turn the ball into fuch 

 a pofition that the plane of the glafs may be, as nearly as 

 you can judge, parallel to the equator; and then incline 

 this plane, in the diredtion of the meridian pafling 

 through the fun, till the fun be about 45" above it. 



Now take your odlant, and by the fore-obfervation, 

 bring one of the limbs of the fun's image, feen by a 

 double refledtion from the fpecula of the inftrument, 

 exadly into contact with the image of the fame limb, 

 feen by a fingle refledlion from the lurface of the glafs 

 plane, and read off the angle pointed out by the index, 

 immediately turn round your inftrument, and bring the 

 fame limbs into contadl by the back obfervation. If the 

 angle now pointed out by the index be exadlly the fup- 

 plement (to 180°) of the former angle, the horizon-glafs 

 for the back-obfervation will be truly adjufted, or ex- 

 adlly at right-angles with the horizon-glafs for the fore- 

 obfervation : But if thefe two be not equal, then take 

 half their difference, which will be the corredion or in- 

 4 dex- 



