388 Special Reports 



All the materials except the marble, the glass, and the copper wire were tested by 

 the mechanicians in the same way in which they are tested for the other instruments 

 constructed in this laboratory. The substance is brought within a few millimeters of the 

 north pole of one of the standard magnetometer magnets, on the west side, and the deflec- 

 tion observed on approach and removal. The deflection can be read to one-fifth minute. 

 Except the telescope, too far away from the magnet and coils to have an appreciable 

 effect, the casting supporting the magnetometer box, and some of the materials in the 

 Edelmann base, the material was found excellent, producing either no appreciable 

 deflection, or no deflection greater than one-fifth minute. The Edelmann base, the 

 copper wire in the coils, the glass, the marble, and the magnetometer and its adjuncts, 

 except the telescope and scale and the small parts holding them, were tested with an 

 astatic magnetometer, whose magnet systems were held normal to the meridian by two 

 small control magnets, the one very small and north of the magnet systems with its 

 axis passing through the meridian intersecting their centers, and the other with its axis 

 in the prime vertical passing through the systems. Both control magnets were near 

 or below the level of the lower magnet system, and did not affect greatly the magnitude 

 of the N-S component of H south of the system. Longitudinal motion of the first magnet 

 altered the zero of the instrument, and longitudinal motion of the second altered the 

 sensitivity, which was determined with a small pair of Helmholtz coils. The sensitivity 

 being determined, the fractional change in H produced by bringing a substance up to 

 within a few millimeters of the south side of the lower system could readily be found. In 

 this way it was found that no one of the elements of construction except the base produced 

 an alteration in H greater than a few parts in 10 6 . Marble and paraffin are, of course, 

 already known to be diamagnetic, with negligible susceptibility, and the copper castings 

 were found to be so nearly pure as to be diamagnetic. The copper wire and the other 

 pieces of metal above the base were all found to be diamagnetic or slightly magnetic ex- 

 cept the one immediately beneath the copper box and the clamp holding the telescope 

 arm. These pieces were about equally bad, and the worst of the metals, producing an 

 alteration in H of about 1 part in 30,000. In the use of the instrument, however, they 

 are so remote from the magnet that their effect is entirely negligible less than 1 

 part in 10 5 . 



To test the tripod as a whole it was placed on a table which could be rapidly moved 

 underneath the shelf on which the magnetometer was mounted, with the top of the 

 tripod about 10 cm. below the lower magnet system, the upper system being 10 cm. or 

 more higher. With the tripod and circles in different azimuths the scale readings were 

 noted when the tripod was placed centrally beneath the magnetometer and when it 

 was moved to a distance. The effect of the tripod on H at the lower magnet was in no 

 case greater than about 2 X 10" 6 . It was by a somewhat similar process that the last 

 statement in the foregoing paragraph was substantiated. 



Finally, the whole instrument, aside from the coil and its spool and with the exception 

 of the copper plug and other fittings to the cylindrical chamber of the magnetometer box, 

 the screws holding the marble spool in place, a few other screws, and the telescope and its 

 adjuncts, was tested with an induction balance. The coil of the instrument formed one of 

 the primaries, and two cylindrical coils in series, one on each side of the magnetometer box, 

 the corresponding secondary. Each of the coils had a breadth of 3 cm. and internal and 

 external diameters of 4 and 8 cm. and was clamped to the spool with the inner face only a 

 few millimeters from the adjacent face of the copper box. When the spool was placed in 

 position on the saddle, with the torsion tube, copper box, and castings A and C (Figure C, 

 Plate 9) removed, or when it was removed to a distance, with all necessary precautions 

 taken, or when, with the coil in position in the saddle, the tube, box, and castings were 

 removed or placed in position, there was no change in the mutual inductance, tested with 

 reversal of current and a ballistic galvanometer, greater than about 1 part in 35,000. 



