Remarks on J-Valker's new Transit Instniment. 289 



eye would be better able to judge when the two small seg- 

 ments are equal than when the whole circle is divided into 

 two equal parts. I am persuaded that the mean of five ob- 

 servations taken hy this method would determine the time 

 of the sun's transit over the meridian to less than a second, 

 with a gnomon only three feet high erected upon a stone 

 three feet in length. 



To illustrate this niode of observing by an 

 Example, 



If fig. 5. (Plate VII.) represent the solar image, and the 

 distance between th^ two lines a and « be = lOS", the two 

 segments will be each = lo" when the sun is in the equator, 

 as he then tai<es about 12S" in passing over a space equal to 

 his own diameter. Then, if an observation be taken only 

 V too soon, the two segments will be to each other in the 

 ratio of 9 to II, which may be too great a difference to pass 

 unobserved. 



When five observations are to be taken by this method, it 

 will be necessary to have seven lines drawn at equal distances ; 

 but the first observation must not be taken until the centre 

 of the solar image be upon the second line, at which 

 tinjc the first and third lines will cut off two equal seg- 

 ments ; when the centre comes to the third line, the se- 

 cond and fourth lines v;ill cut off two equal segments; 

 and the third, fourth, and fifth observations must be taken 

 in tiic same manner. If these lines he drawn at such di- 

 stances that the centre of the solar image may take about 

 54 seconds in passing from one line to another when the 

 sun is near the equator, the five observations may be taken 

 in less time than four minutes. 



This instrument, though erected at a small expense, is 

 infinitely more convenient for keeping the rate of a clock 

 or a chronometer than either Hadley's sextant or the astro- 

 non)ic?.l quadrant, and perhaps superior to eith«;r of them 

 in point of ac(;uracy. — I am, sir. 



Your most humble servant, 



i.ynn, April 5, 18D7. Ez, W.ALKER. 



N. B. Tlie letters no are omitted in Plate III. vol. xxvii. 

 JirraU7a.—0r. page i»a, line IS, for clj;cts read dji.i.1. 



T 2 L. Ofi 



