114 



Scientific Proceedings, Royal Dublin Society. 



free, it did not return to zero as in the other four wires, but was from 5" to 

 10° out. The positions therefore of these points on the curve is what one 



Q IS 30 4S 



Cross-sectional Area of Wire in sq. cms. x 10"'. 

 Fig. 7. 



would expect, for if the outer layers on the wires get a permanent shear, or 

 are taken beyond the elastic limit, the molecules of the iron in these layers 

 would contribute little or nothing to the circular magnetism or area of the 

 cyclic curve. 



la order now to find the effect of varying the circular magnetism on a 

 given wire with different loads, the ballistic galvanometer was taken out and 

 the circuit re-arranged. 



In series with the wire under test were put a storage-cell, ammeter, 

 variable resistance, reversing key, and plug key. A No. 16 iron wire was 

 taken, and carefully annealed, and hung in the Earth's vertical magnetic 

 field with a given weight on the free end as in fig. 1. On the lower end of 

 the wire a small plane mirror was fixed, and a lamp and scale suitably 

 placed ; by this means the twist of the end of the wire could be measured. 



Now seeing that when we twist an iron wire which is placed in a 

 magnetic field a momentary current is produced, it follows that if we send 

 an electric current through the wire the free end of it will twist. This is the 

 Wiedemann effect. 



The circuit being arranged as indicated above, cui'rents were sent 

 through the wire in eight different steps, from zero to 2'06 amperes, and the 

 twist or deflections on the scale read off in each case. The current was then 

 diminished by the same steps to 0-12 amperes, and the circuit broken, the 



