INSTRUMENTS AND KECORDS. 



75 



5. The Declination magnetograph remained at a fixed distance from the recording drum and its sus- 

 pension was unaltered, so the scule value was accepted as constant. The light from the lamp, emerging 

 from a narrow slit, traverses a slightly convex lens forming a window to the magnet box, is reflected from 

 a plane mirror carried by the magnet, passes a second time through the lens window, and eventually falla 

 on the photographic paper, after passing through a hemi-cylindrical lens a little in front of the paper. 

 The thickness of the hcmi-cylindriciil lens, approximately 0'5 cm., and the focal length of the lens window, 

 about 120 cm., exercise a slight effect on the scale value; but for practical purposes it depends essentially 

 on the distance of the plane mirror from the photographic paper. The scale value calculated from the 

 optical conditions in the Antarctic was 1 mm. = l' - 43. This requires, however, a small correction for the 

 torsion of the suspension, a very fine quartz fibre. This suspension returned safely to Europe, and 

 independent observations by Mr. BERNACCHI and the Kew staff in which the torsion head was twisted 

 through 180 made the torsion correction very approximately 4 per cent. The scale value for the 

 Antarctic curves was thus taken as 1 mm. = l'-50. 



6. The scale values of the Horizontal- and Vertical-Force magnetographs were determined from time to 

 time in the Antarctic by means of one or other of the collimator magnets 25A and 25b supplied with the 

 unifilar magnetometer. In the case of the Horizontal-Force magnetograph, the magnet was placed 

 horizontally in front of the recording drum in the magnetic Meridian. A paper being on the drum, the 

 position of the spot of light was recorded photographically with the magnet as originally laid down and 

 when turned end for end. The whole operation was repeated at least once on each occasion. The double 

 deflection, i.e., the distance between the two deflected positions of the spot of light, answers to a change of 

 very approximately 4m/r 3 in the Horizontal Force, r being the distance between the centres of the 

 deflecting and the deflected magnets (usually at least 150 cms.), and m the moment of the magnet at the time. 

 The change in m was small and slow, so that a sufficiently approximate value was derivable from the 

 ordinary absolute observations. The procedure in the case of the Vertical Force was practically the same 

 except that the collimator magnet was hold vertically over the Vertical-Force magnet. The scale values 

 thus found and the values actually employed are recorded in Table I. 



TABLE I. Scale Values. 

 Value of 1 cm. of ordinate in terms of ly (ly = -00001 C.G.S.). 



L 2 



