THÉ DISTURBANCE OF ISOMAGNETICS. 



165 



other at the Ijack, These tliermometers were compared with the 

 standard and corrected. 



Daring the former survey, the magnet remained straight and came 

 into good contact with the stops and therefore the distances, i\ and 

 7*2, of tlie magnet from the magnetometer magnet were known from 

 the measurements of the distances a a, h h', and the length of the 

 magnet. 13ut as the magnet had hecome decidedly curved, it would 

 no longer come into good contact with the stops. Generally, only 

 one extremity of the diagonal of the square cross-section wîis in con- 

 tact, while there was a slight gap at the other end. This change of 

 contact necessitated the redetermination of the distances of the centre of 

 the magnet in positions a, h, a\ h' , The actual distances of the centre 

 of the magnet from the magnetometer magnet can only he determined 

 by placing the magnet in the groove in the same way as in the Jictual 

 deflection experiments. 



The deflection rod was placed on a plane plank, and the lower 

 side of the metre scale was brouglit in contact with the upper face of 

 the rod. To guard against any change of position of the metre scale 

 and the rod, the latter was gently clamped to the plank at its middle, 

 and any relative moti(}n which might take place between the two 

 during measurement was noted by a microscope of high magnifying 

 power. No relative motion of the scale and rod was observed during 

 the operation. The magnet being placed in the usual way in the 

 groove, the edge of the magnet adjacent to the stop was sighted, and the, 

 readings of the scale and micrometer noted in the four positions a, h. 

 a\ h'. The magnet was then reversed, and the process repeated. 

 The rod beim?- next reversed, the readinj^'s of the other edo;e of the 

 magnet were noted in the same position as before. For each reading 

 of the micrometer, four thermometers were read, two being placed on 

 the scale and the others close to thé rod. The mean of the observed 



