169 



pulleys attached directly to the screw, we know the error 

 would be still less. From this fact, we were led to surmise, 

 that difference of declination could easily be read to the 

 tenth of a minute, from a screw head used for giving slow 

 motion to the Telescope. 



In thinking on this subject, T conjectured that if a cylin- 

 der were attached to this screw, and a pen be made to 



Fig. 1. 



move over it with a uniform velocity in the direction of 

 its length, we could readily record both Right Ascension 



