OP THK UNITEU STATES. 337 



to receive the transit of tlie star, or both limbs oi the sun, at 

 all the horizontal wires. In this situation the level is adjust- 

 e(J with ease, sulTicient time being given, by the placing of 

 the circle ; and the transit of the star, or both limbs of the 

 sun, at eacii of the thiee wires is observed. This can be 

 done with tlie greatest nicety, tlie instrument being at rest 

 and all tlie observer's attention being directed to the time of 

 contact which is rapid; the observer calling out J^ull, at 

 each transit, the secretary being attentive to tlie chronometer 

 or clock, can note the time with the greatest accuracy ; and 

 the transit of the sun's centre is determined I)y six observa- 

 tions. 



The circle being now brought into its second position, by 

 a semi-revolution on its vertical axis, the level will remain 

 adjusted, if the instrument is well levelled. If it require any 

 correction, this must be made by the circle screw rn, as the 

 position of the level upon the circle must be preserved. 



The telescope is then undamped at ]), and again placed 

 so as to receive the transits as before. These are observed 

 and the time noted, as has been done in the lirst position. 



This observation being made, the verniers of tlic front te- 

 lescope are again read, and give the double zcnitli distance, 

 or rather the sum of the two zenith distances at the times of 

 the observed transits. The half of this sum is to be consi- 

 dered as corresponding to the mean time between the two 

 transits. 



Such an observation, when taken near the prime vertical, 

 will not exceed the time allowed for taking an arithmetical 

 mean, and will be the result of six or twelve observations of 

 time. 



If the transit of the sun is supposed to take too much time 

 to observe both liml)s on each side, tlie antecedent may l)c 

 taken in the lirst position, and the consequent in the second. 

 There being then but half the number of times obtained, the 

 observation may be repeated after the reading, and a mean 

 of two such observations taken, as in any other case, calcu- 

 lating each separate. 



S 



