26 Land Magnetic Observations, 1905-10 



where Z is the vertical intensity of the Earth's field at the place of observation and 

 m the magnetic moment of the magnet whose induction coefficient is being obtained. 

 The quantity, ju = mh, may be derived directly by the formula: 



, mi(a Oo) 

 2Zax 



Plate 6, Fig. 2, shows the apparatus used in determining the induction coefficient 

 by Lamont's method. The magnet whose induction coefficient is to be determined 

 is used to deflect a suspended magnet mounted as usual in the magnetometer, 

 the deflecting magnet, however, having its length in a vertical direction at right 

 angles to and some distance above or below the horizontal plane through the 

 suspended magnet. The arm carrying the magnet being tested may be quickly 

 rotated through 180 vertically; suitable arrangements are made for adjusting the 

 center of rotation to the height of the suspended magnet. Deflection observations 

 are made (1) with the north end of the magnet up and the magnet up, then (2) north 

 end of the magnet down and the magnet down; the magnet is then turned about so 

 that (3) the north end is up and the magnet still down, and finally the arm is 

 rotated through 180 so that (4) the north end of the magnet is down while the 

 magnet itself is up; these operations are then repeated in the reverse order. The 

 change in the resulting deflection angles measures the alteration in the magnetic 

 moment due to the induction of the Earth's vertical intensity, Z. The induction 

 coefficient, h, is computed by the formula: 



, _ tan |(m2 Ui) 



H tan /tan \{Ui + iiy) 

 u-i and i being the mean deflection angles obtained from the positions 2 and 4, 

 and 1 and 3 respectively. Such observations, in order to eliminate accidental 

 errors, are made at three or four deflection distances, both horizontal and vertical. 

 Except when there are a great many magnets to be tested at the same time Lamont's 

 method is on the whole more convenient. 



For the determination of the distribution coefficients P and Q, observations 



of deflections are made at three distances, usually twelve or more complete sets 



being obtained. The outstanding differences in the values of C'/sin u for the 



three distances, where 



2 



C = 



r'{^+f 



r 



are assumed due to the effects represented by the distribution coefficients P and 

 Q. The resulting equations may be solved by the following substitutions, the dis- 

 tances used being ri, r^, and rz (fi being the short distance) : 



R_l 1 R'-l-l 5--I_i B'"-^-- 



( C2 C'l \ ^^ ( C3 ^2 \ ,^/ 



Vsin 2 sin Ui) Vsin M3 sin U2/ 



_ A'B" - A B'" __ A'B - AB ' p == p'^ o = o'E 



~ B'B" - BB'" ^ B'B" - BB'" m ^ ^ m 



