A NEW CURRENT WEIGHER, ETC. 4/5 



To allow of free access to the balance there are two sliding sashes, S', S', fig. 6, both 

 at the front and back of the case, and the ends have hinged doors, D, opening out- 

 wards. The middle portion of the case carrying the microscope and the corresponding 

 piece at the back can also be removed. It is thus possible to make any adjustment 

 required with comparative ease. Fig. 2, Plate 7, shows a view of the balance with 

 sides of the case taken away. 



Although it is not essential that the arms of the balance should be of equal length 

 in a current weigher used in the manner described on p. 526, it was thought desirable 

 to determine the ratio of their lengths. Employing weights of 50 grammes, it was 

 found that 



length of left arm -5- length of right arm = 1 "00001,, 



so a very close approximation to equality exists. 



SECTION 5. MAGNETIC TESTS. 



As it is of considerable importance that the permeability of all parts of the current 

 weigher be practically unity, magnetic tests were made on the materials employed. 



Before the phosphor-bronze support for the balance was cast, Sir ANDREW NOBLE 

 forwarded to the Central Technical College in September, 1899, a bar of the alloy it 

 was intended to use, and careful experiments were made on the material. Tested by 

 a very sensitive magnetometer the bar showed no magnetic property. An induction 

 balance having two primary coils in series, and two secondaries in series opposing, with 

 a sensitive moving-coil galvanometer in the same circuit, was therefore set up. One ot 

 the induction coils was in the form of a solenoid 2'4 centims. diameter and 36 centims. 

 long, wound with 457 turns of No. 18 S.W.G. wire as primary and 1600 turns of 

 No. 34 S.W.G. as secondary. The other half of the balance was formed of two 

 separate coils whose relative position could be varied continuously until their mutual 

 induction exactly balanced that of the solenoid windings when the core was of air. 

 By shunting a known fraction (joW) f the current from the primary of the second 

 pair the swing obtained on the scale of the galvanometer gave a measure of the 

 sensitiveness of the arrangement ; this was sufficient to show a change of 1 part in 

 30,000. 



On removing the above-mentioned shunt and inserting the phosphor-bronze rod 

 (2 centims. diameter and 30 centims. long) into the solenoid, a quick jerk of the 

 galvanometer spot was olwerved on starting and stopping the primary current, and a 

 rapid return to zero. The direction of the kick was such as would be produced by a 

 permeability less than unity ; the effect, however, was traced to be mainly due to 

 eddy currents in the rod, and was nearly neutralized by putting a tertiary coil, with a 

 resistance box in series with it, in proximity to the second pair of coils. The 

 resistance in the tertiary circuit could be so adjusted that the movement of the spot 



3 P 2 



