254' Mr. J. Nixon on the Trigonometrical Height 



ground apparently quite firm, a certain angle was satisfactorily 

 repeated to the fifteenth time, when it was discovered that my 

 moving two or three feet to the right caused the telescope to 

 swerve in azimuth. On repeating the measurement at the 

 foundations of the old building, the previous one proved to be 

 nearly 15'^ in excess*. 



The Ys being constructed of equal height, (as was proved 

 by the reversed telescope pointing at the previous mark on 

 turning the circle half round in azimuth,) the line of collima- 

 tion was rendered parallel to the (divided) plate by adjusting 

 it to bisect the same distant mark during a revolution of the 

 telescope within its Ys. One of the cross wires being set 

 parallel to the plate, (in which case both extremities bisect 

 the same object as the plate revolves,) a minute dot, adhering 

 to it a little to the left of the vertical one, was advantageously 

 substituted for their point of intersection. 



The graduated plate is got into the plane of the two objects 

 by the screws of the parallel plates. Unfortunately the socket 

 yielded to thefo7rible action of the screws, which increased 

 at every adjustment the distance between the parallel plates 

 and produced a simultaneous movement in azimuth. On 

 finishing an adjustment, in order to avoid any subsequent 

 deviation of the circle in plane or possibly in azimuth, the 

 screws were left equally tight. The process of observation 

 may be briefly described with reference to the figure. Having 

 fixed the clamp I, turn the pinion H until the telescope bi- 

 sects the signal jt. Fix the clamp J, and having registered 

 the reading, unscrew the clamp 1. Move the pinion G to 

 bring the telescope to bisect the signal 7/. Then fix the clamp 

 I, and having obtained the reading, unscrew the clamp J, 

 and turn the pinion H until the telescope points again at x. 

 Lastly, fix J, and having unscrewed I, move the pinion G 

 until the signal y becomes bisected the second time. Fix I, 

 and register the reading, which will differ from the one first 

 obtained by double the angle required. After this manner 

 any other multiple of an angle may be measured. Bisections 

 made with rack-work and pinion are, from the interfering ac- 

 tion of the clamp, more difficult and tedious to effect than 

 by a tangent screw, (of which two are used in repeating with 

 a theodolite.) but unquestionably they are more permanent fi 

 Several angles measured on the theodolite would exhibit at 



♦ A circle with two telescopes would have been equally liable to the 

 error. 



t Mr. Lloj'd's superb level (by Gary) would not retain the adjustments 

 of its bubble when effected by an endless screw. 



