Bifilar or Horizontal Force Magnetometer. 
XXXI 
coefficient of the temperature correction for the varying magnetic moment of the 
bar or the value of for 1° Fahr., e and e the coefficients of expansion for the 
brass of the grooved wheel and silver of the wires. 
36. The observations in this volume are given in scale divisions, and are cor- 
rected by the coefficient — +-=?> q being the total temperature coefficient; the 
Gr COX V /c 
A X 
abstracts are then obtained from the formula = n a cot v where n is the number 
of scale divisions corrected for temperature as above. 
37. The following are the adjustments and values of the constants. 
The angular value of one scale division of the bifilar magnetometer — 1/-1223 
(See Table 10, Introduction, 1841-2) ; increasing readings indicate increasing force. 
The value of q, the total coefficient of the temperature correction, = 0-000247 (71.) 
38. It having been suspected that the zero of the scale (the scale reading when 
w=90°) had altered in some way since last adjustment on Oct. 20. 1841 (Introduc- 
tion, 1841-2, p. xxviii.), the following observations were made 
April 27 d 2 h 39 m . Bifilar scale reading 196-9. 2 h 44 m . Bifilar scale reading 
196*6 ; the magnet was then removed carefully, and the equivalent brass weight 
substituted. The torsion circle was then turned from the reading vernier A 289° 10' 
to A 358° 16', or through 69° 6', the previous value of v (Introduction, 1841-2, 
p. xxviii.), when the scale should have read 150, instead of which it read 212*2. As 
it seemed possible that this difference might be due to pressure exerted on the wires 
in withdrawing the magnet and substituting the weight, the magnet was again in- 
serted in the stirrup, the weight being removed, and the torsion circle turned till it 
read A 289° 10', when the scale read 297. It was therefore evident that the pre- 
vious difference might be due in some way to the manipulation it was necessary, 
however, to go through the adjustment anew. 
After several trials the scale reading was found to read the same whether the 
brass weight or magnet was suspended, when the torsion circle read A 87° 30'. The 
weight being suspended, the torsion circle was turned 90°, when it read A 357° 30'. 
The collimator was then turned by its independent motion till the scale read 200. 
The magnet being again suspended, the scale reading was found 200 when the tor- 
sion circle was turned 69° 45', it then read A 287° 45'. 
April 28 d 0 h 20 m . As the zero of the scale, 200, was taken too high, the adjust- 
ment was again performed. 
Bifilar scale reading 200. The magnet being withdrawn, the brass weight sus- 
pended, and the arms of the torsion circle turned from A 287° 45' to A 357° 30', the 
scale reading was found 199, so that the wires had not been affected in the previous 
adjustment. The collimator was then turned till the scale read 148, this being the 
mean of several readings; the magnet was again suspended, and the torsion circle 
