326 SURVEY OF THE COAST 



is very possible, the circle should be used in the two invert- 

 ed positions which the semicircles admit. 



Peculiar Method of Observing Time with the Repeating 



Circle. 



In the observations for tlie determination of time, this ele- 

 ment itself is the principal object of research, and may be 

 considered as the most difficult to obtain accurate. The pro- 

 per adjustment of the level in an observation of such a tran- 

 sitory nature, and during which the circle may move con- 

 siderably in tiie vertical by pointing, requires great dexterity 

 in the assistant ; and since the level itself oscillates in passing 

 to equilibrium, it is often very difficult to be sure of its po- 

 sition. 



From one complete observation with the circle in its two 

 positions and the indiscriminate mean of the four readings, 

 all the main corrections of the observation are obtained. Suc- 

 ceeding observations serve only to augment the probability 

 of accuracy, and to correct accidental errors of division, 

 which, as already stated, may be considered a minimum in 

 the instruments of Mr. Troughton. 



A method satisfying the two first mentioned desiderata 

 will therefore secure more accuracy in this kind of observa- 

 tion than the usual mode of repeating. IVly peculiar situa- 

 tion, with assistants entirely unacquainted with observations, 

 joined to these considerations, led me to devise the follow- 

 ing method of observing time, which I have ever since prac- 

 tised, and which Dr. Tiarks, the British astronomer for the 

 boundary line at Canada, adopted also in our common works 

 there. 



The instrument being well levelled and the position of the 

 front telescope read off as usual, the circle is placed so as 



