J4Q METHOD OF TAKING TRANSIT OBSERVATIONS. 



thq star f ror n the wire at the two beats of the clock, pra-v 

 poses another, which he thinks superior. This consists in 

 noting the time when the centre of the star comes to one 

 side of the wire; which, lie observes, is a real line, and not 

 as in Bradley's, a line drawn by the strength of imagination 

 down the middle of the wire, parallel to the sides. I have 

 tried both these methods with nearly the same success, and 

 must confess that, after all, I am very much at a loss to 

 conjecture, how the fractional part of a second can be esti- 

 mated in either of these ways to that nicety it appears to be 

 done. In the observations made at Greenwich I observe, 

 the time of a star's passing the meridian is always expressed 

 to the hundredth part of a second. How this extreme pre- 

 cision is obtained, as I am at a loss to conjecture, I shall be 

 obliged to you, or any of your correspondents, to inform 

 we, 



I am, Sir, 



Your obedient servant, 



J. G . 



REPLY. 



Mfithodoftik- It is certainly not difficult to observe to tenth parts of a 

 z transit ob- ^econd ; and of this my correspondent will easily satisfy 

 himself by trial with a common watch of five beats in a se- 

 cond. A phenomenon, as for example, the transit of a 

 star, may take place in any one of the five beats, or bat- 

 tween any two of them. If the observer repeat the words 

 (either mentally or otherwise) One, one, one, one, one— 

 Two, two, two, two, two — Three, three, three, three, three, 

 &c, at each beat of his seconds clock, the word in Italic 

 at the very beat, he will be enabled to mark the fractions 

 of seconds with great precision. Musicians, in the rapid 

 execution of prestissimo movements, divide the second still 

 lower. As to the hundredth parts of seconds, though it 

 might by some expedients be practicable to observe them, 

 this is not implied in Astronomical Tables. They are 

 almost always the reswlts of means taken between a number 

 o.f observations; and the secoud decimal may be considereo} 



as 



HI 



