MAGNETIC OBSERVATIONS. 43 



suspended in the magnet-box; and the telescope having been pointed so that its 

 vertical wire cut the magnet scale approximately at the magnetic axis, the hori- 

 zontal limb of the theodolite was firmly clamped. Then, 1. By quickly approach- 

 ing and withdrawing the magnetised screw-driver the magnet was caused to vibrate 

 horizontally through an arc extending to about twenty scale divisions on each side 

 of the magnetic axis that is, through a total arc of about 1 34'. The semi-arc 

 of vibration being only 4?', no correction to the observed time of vibration was 

 ever required on that account. 2. My assistant having taken the chronometer, I 

 placed my eye to the telescope, and at the instant the 80th division of the scale 

 (which was very near the magnetic axis) crgssed the vertical wire I cried "time," 

 and my assistant noted the minute, second, and fraction of a second indicated by 

 the chronometer. Still keeping my eye at the telescope, I counted the transits of 

 the 80th division over the wire, calling the one at which time was noted 0, the 

 next 1, the next 2, and so on up to the 10th, when I again cried "time," and my 

 assistant once more noted the minute, second, and fraction of a second indicated 

 by the chronometer. The difference of these two chronometer times gave a value 

 for the time of ten vibrations of the magnet which was correct within about half a 

 second. However, to guard against mistakes, the process was always repeated a 

 second or third time. 3. The temperature indicated by the thermometer in the 

 magnet-box was noted; and then putting my eye to the telescope, I read the scale 

 at the instant the magnet attained the eastern extremity, and again when it attained 

 the western extremity, of its arc of vibration. These were the "extreme scale 

 readings." 4. The chronometer employed was a pocket one, beating five times 

 in two seconds. Taking it in my hand, I commenced counting its beats at some 

 multiple of ten seconds. Then, holding it to my ear and still mentally count- 

 ing the beats, I put my eye to the telescope and noted the beat, and fraction of a 

 beat, at which the 80th scale division crossed the vertical wire. For example, 

 suppose the beat was taken up at the instant the chronometer indicated 10 h 2 m 10 s , 

 and counting the first succeeding beat 1, the next 2, and so on, suppose that the 

 80th division crossed the wire exactly at the 14th beat. Then, as 14.0 beats are 

 equal to 5.6 seconds, the time of transit of the 80th scale division was 10 h 2 m 15 3 .6. 

 The time of transit thus obtained was recorded as the vibration. Adding to it 

 the time of making ten vibrations before determined the approximate time when 

 the 10th vibration would be completed became known. Taking up the beat of the 

 chronometer at the nearest even ten seconds before that time, I put my eye to the 

 telescope and observed the time of transit of the 80th division at the completion 

 of the 10th vibration. In the same manner the time of completing the 20th, 30th, 

 40th, 50th, 100th, 150th, 160th, 170th, 180th, 190th, and 200th vibration was 

 observed. Subtracting the time of completing the vibration from the 150th, the 

 10th from the 160th, &c., there result six values of the time of making one hundred 

 and fifty vibrations, from the mean of which a very accurate value of the time of 

 making one vibration is obtained. It will not escape notice that when observing 

 in the manner just described there is no risk of making a mistake of one vibration, 

 because the magnet must, at all subsequent transits, be moving in the same direc- 

 tion as at the first transit, while in order to make a mistake of one vibration it 



