T J? Period of magnet alone; T 2 , 

 A, Moment of inertia. 



Feb. 29 21-918 



March 1 (!) .... 21-870 



(2) ... 21-906 



March 2 21*893 



Magnet I. 



A. 

 13392 

 13337 

 13379 

 13373 



T 2 . 

 38-759 



38-730 

 38-753 

 38*735 



Mean 



13370 





442 Mr. R. H. M. Bosanquet on the Horizontal 



that obtained by direct measure. The magnets were distorted 

 by hardening and cannot have been homogeneous, so that 

 the direct measure was untrustworthy. The brass bar was 

 put in the place of each of the magnets in turn, and the times 

 of vibration of each magnet observed with and without the 

 brass bar. This is the least satisfactory part of the whole 

 process. It is impossible to get rid of disturbance due to 

 temperature and changes of H, and these affect the calcula- 

 tion considerably. The following experiments were used to 

 determine the moments of inertia. 



of magnet and bar. 



Magnet II. 

 T r A. T 2 . 



21-780 13374 38*538 



21-815 13316 38-656 



21-810 13309 38-654 



21-846 13337 38689 



13334 



These were taken at a temperature averaging about 60°. 



The vibrations of the magnets were observed by placing 

 a small plane mirror on the brass-wire stem by which the 

 magnets are supported. A circular scale of degrees is placed 

 in front of it, and an upright wire in front of the scale 

 adjusted to the position of rest. Then the arc can be read off 

 in degrees and adjusted to a suitable amount, and the obser- 

 vations of the passage of the wire over a black silk cross on 

 the mirror are convenient to make. The clock employed 

 during these observations had a losing rate of about 8 seconds 

 per day. It was checked by comparison with a good clock 

 which has a losing rate of *66 seconds per day, as determined 

 by comparison from the Radcliffe Observatory. 



The readings generally included about an hour, so that the 

 five places in the period were pretty well determined. An 

 inconvenience of this long duration is, however, that changes 

 of temperature or changes of H sometimes set in which 

 rendered the different parts of the time nonconformable. In 

 these cases the period deduced from the first ten minutes or 

 half-hour would not correspond with the whole or with that 

 deduced from the last half-hour. Such observations had to 

 be rejected. 



The observations were made by the method indicated by 

 Maxwell : six observations of passages to the right and five 

 to the left were used to form each normal time ; one of these 

 sets was usually made at starting, and others after ten 

 minutes, after half an hour, and an hour. No doubt it is an 



