598 



I3IRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



116. A new glass box was constructed to contain more than 70 litres, in order that all discharge- 

 experiments might be made with a terrella so strongly magnetised that the cathode-ray system round 

 it would be similar to the corpuscular-ray system round the earth. The thickness of the plate-glass 

 sheets was 22 mm.; and in order to guard against the great external pressure, the sheets forming the 

 ends of the box were specially strengthened. The sheets that were perforated were double. The 

 internal dimensions of the box were 36 X 36 X 55 cm. The terrella employed was 8 centimetres in 

 diameter, and was constructed with the object of procuring more than 4500 lines of force per centimetre 

 across the poles when the strongest magnetising current was employed. The iron core was 3 cm. in 

 diameter, and was closely wound round with well-insulated layers of copper wire, of which the total 



Fig. 219. 



resistance was 2.6 JL'. The wire could, without injury, be charged for a few seconds with 40 amperes 

 thereby imparting to the terrella an amount of energy equal to between 5 and 6 horse-power. 



The magnetic moment M for 10 amperes was found to be 61300. At a distance of 4.5mm. from 

 the terrella, immediately above the- pole, the number of lines of force with 10, 15 and 30 amperes' 

 magnetisation was respectively 2075, 2760 and 4200. At a distance of 7.5 mm. from the pole of the 

 terrella the measurements were H = 1647, 2460 and 3280 with 10, 15 and 20 amperes respectively of 

 magnetic current. These measurements were taken with a Leduc apparatus. To ascertain if it were 

 correct, this apparatus was compared with a Lenard's bismuth spiral which gave the following correspon- 

 ding sets of values: 



H by Lenard spiral 575 5600 4800 3550 



H by Leduc apparatus .... 5950 5650 5040 3720 



As will be seen, the respective values correspond fairly well and the records of the Leduc 

 apparatus must thus be considered reliable. From this it must be supposed that immediately above 



