760 



SCIENCE. 



[N. S. Vol. II. No. 49. 



rent in the coil and K its factor. KC can 

 be found empirically by assuming values 

 for d and H, which will be convenient and 

 leave the magnetometer at a suitable dis- 

 tance from the scale S. Of course if it is 

 preferred one can use a millimeter scale and 

 read the tangents and reduce each time. 

 The operation with the apparatus is as 

 follows : 



the horizontal intensity found for the fidu- 

 cial locality. The coil B is then slid back- 

 ward or forward until the mirror normal 

 falls upon the cross of the sight. Every- 

 thing is then left in position and the vari- 

 ometer taken to the, second station; the 

 sight being adjusted normal to the new de- 

 flection of the mirror, the reading will be 

 the intensity at that place. Thus the in- 



Before the current in B is started, the 

 sight is put at zero and the instrument ori- 

 ented at the standard locality as described 

 for Fig. 1. This oriented position should 

 be so marked that the variometer can be 

 easily and quickly returned to it. Simi- 

 larly the oriented positions at the various 

 stations should be established before com- 

 mencing the observations. 



The current is started and allowed to flow 

 till fairly constant. When observations 

 are to be made, the sight is placed at the 

 point upon the scale reading the same as 



strument gives the horizontal component 

 directlj\ Naturally one must always take 

 the precaution to commutate the current, 

 and take frequent check readings at the 

 fiducial location. 



In this way observations about a room or 

 building can be conveniently made and re- 

 sults obtaiued whose errors are certainly 

 not greater than the daily and other fluctu- 

 ations of intensitJ^ Naturally by increas- 

 ing the distance from the mirror to the 

 scale and the j)i-ecautious for constancy of 

 current, one can obtain more acurate re- 



