102 ON THE REPEATING REFLECTING CIRCLE. 



contact, which I remark. I then move forward the index 90°, 

 and thus I repeat the operation successively through each qua- 

 drant of the 720 9 . Should they appear quite coincident during 

 the whole revolution, it is a proof not only that the two mirrors 

 are parallel, but that the plane of the circle is a true one, 

 which rarely happens. Should I, however, find, that the reflec- 

 ted sun appears sometimes to the right and sometimes to the 

 left of the one seen direct during this operation, it leads me to 

 suspect that one of two causes of error exists ; either that the 

 plane of the circle is not just, or that the two mirrors are not 

 parallel. This important point I determine, by examining the 

 adjustment of the great mirror, which, if found correct, I then 

 ascribe to the inaccuracy of the plane of the instrument, 

 which deviation I distribute equally throughout the whole 

 circle. 



The next principal adjustment, is to ascertain if the tele- 

 scope be parallel to the plane of the circle, which is done ei- 

 ther by a small proof telescope applied to the circle, when in a 

 horizontal position ; and, if the object seen at some distance 

 appear in the centre of that telescope, and the one attached to 

 the circle, it is a-proof of its correctness. There are generally 

 two small sights sent with the circle, to ascertain this adjust- 

 ment. 



Having carefully examined these, and corrected any error, 

 if it exists, before each day's observations, we may rest assured 

 that the results will be satisfactory, if no error be committed 

 in determining the time of apparent noon, the observations 

 properly noted, and the calculations rigorously made according 

 to the accompanying formulae. 



Paris, 25th March 1819. 



VI. 



