TRANSACTIONS OF SECTION G. 765 



ing. The water-circulation was effected by two small brass tubes, about 0-75 cm. 

 diameter, entering laterally through the solenoid tube beyond the ebonite end into 

 the spaces between the tubes. 



Combined Galmnometer mid Magnetometer. — This instrument was simply a 

 mirror galvanometer, mounted on levelling screws, and with two coils of copper 

 wire fitted closely on each side of the needle. The instrument was made fairly 

 dead beat by employing a large thin aluminium vane, just fitting the needle 

 chamber, to carry the mirror and needle. The resistance of the coils was found to 

 be 100-3 ohms at 18°-7 C. It was used for measuring currents and resistances by 

 the fall of potential method. The direct effect of the solenoid was balanced in the 

 usual way oy a balancing coil. 



The Mountiiiq for the Iron Wive Specimen.—The iron wire specimen was 

 26-1 cm. long and 6-127 cm. diam. Its ends were fused to copper wires 10-5 cm. 

 long and 0-40 cm. diam., an arrangement intended to conduce to a constant tem- 

 perature throughout the length of the iron wire when the heating current was 

 passed through the circuits. The ends of the copper wire were riveted and tin- 

 soldered to two Jt B. & S. copper rods, which were brought out at both ends of the 

 containing tube. • The rods were at about 6 cm. from their inside ends, and copper 

 springs, 3 strands t 18 B. & S., wound oppositely, were introduced to take up the 

 slack of the specimen and copper wires when heated. The springs fitted the inclosing 

 glass tube fairly well, but better centring was obtained by brass washers soldered 

 on the 6 cm. segments at about 5 cm. apart and turned up to fit the tube. 



Two fine platinum wires, for potential leads, were attached at 15 cm. apart to 

 the iron wire specimen. They were brought out in glass capillary tubes running 

 through diametrically opposite holes in the washers and inside the springs to one 

 end of the glass containing tube. The ends of the capillary tubes were allowed to 

 protrude about 5 cm. beyond the containing tube, and were fused at both ends to 

 the platinum wires. 



The diameter of the platinum leads was 0-005 cm.; total length, 98 cm.; 

 resistance, 8-06 ohms. 



The glass containing tube was made tight at the ends in the following manner. 

 At the end where the capillary tubes were brought out a large brass cup, 2 cm. 

 deep and 2 cm. internal diameter, drilled through the bottom to fit the rod and 

 capillary tubes, was threaded over into position and soldered to the rod. This 

 tube just fitted the inner solenoid tube, and had an internal diameter 

 of about 1-27 cm. It was slipped over the apparatus till its end reached the 

 bottom of the cup. The cup was then filled with melted fusible alloy, and to 

 make tightness doubly sure a mixture of beeswax and resin was run around all the 

 joints. The other end of the glass tube was drawn down so as nearly to fit the 

 copper rod, and a similar but, on account of having to go through the solenoid 

 tube, smaller cup soldered to the rod, after putting a small initial extension of 

 about 2 mm. in the springs. 



The seal was made in the same way as before. The connection to the pump 

 was made by means of a small copper tube entering through the bottom of the 

 brass cup and sweated in with solder. It was bent up at a right angle to facilitate 

 a mercury immersion joint. The length of glass tube between the cups was 

 61 cm., which was about 2 cm. longer than the inner brass tube of the solenoid. _ 



"When the apparatus was in place so that the specimen was symmetrical with 

 respect to the solenoid, the small brass cup was just at the end of the inner solenoid 

 tube. Three brass set screws, ranged symmetrically, were tapped through this end 

 of the solenoid tube and screwed down hard on the cup. At the other end of the 

 solenoid tube a brass binding screw was soldered. This, with a similar screw on 

 the adjacent end of the copper rod, formed the heating current terminals. By 

 making the heating current return around itself in this way, its direct efiect on the 

 magnetometer was minimised. 



A compressed fibre block was screwed to the copper rod near the ends of the 

 capillary tubes, and vertical holes drilled side by side in it, to form mercury cups 

 for the platinum leads. Twin wire leads of approximately equal resistances were 

 brought from the terminals of the standard resistances in the magnetising and 



