98 ON THE REPEATING REFLECTING CIRCLE. 



tudes, and found the results the same. I have also ascertain- 

 ed the accuracy of the instrument from latitudes determined 

 by a sixteen inch repeating circle with a moveable axis, and 

 level of Reichenbach's, with which I had frequently observed 

 at Valenciennes, Blandecques, &c. ; and it is but justice to 

 say, that in these instances, Mr Troughton's circle gave the 

 same results to a second. I beg to be permitted to be a little 

 more particular than I otherwise should, as to the mode I have 

 pursued in calculating and deducing latitudes from this instru- 

 ment, as it is not by any means as -yet generally known in 

 England, although I have no doubt, that when quite under- 

 stood, it will be found to surpass all other instruments of its 

 size, for simplicity and accuracy, and I am desirous that 

 amateurs may profit from the experience I have had for some 

 years of its advantages. 



In order to determine the apparent time by the clock or 

 chronometer, I generally begin two, three, or four hours be- 

 fore noon, to take the three series of repetitions, and from the 

 observed angles I infer the time, as pointed out in my former 

 memoir on the sextant. If the altitudes corresponding to 

 these can be obtained in the afternoon, I prefer equal altitudes, 

 or otherwise the simple repetitions made both before and af- 

 ter noon combined, and the mean taken. 



I generally begin to repeat about 7' before the sun's passage 

 over the meridian, which leaves ample time for the three se- 

 ries, the middle one of which I generally contrive to be divid- 

 ed by noon, by which means the errors are compensated, 

 should any exist in the determination of the time, or in the 

 reductions. 



The third series is also finished about the 7' past noon ; and, 

 although it is not what I would recommend except in extreme 

 cases, I have been induced to begin 20' before noon, and, pro- 

 vided 



