20 



ROYAL SOCIETY OF CANADA 



strip was connected to a galvanometer G, and a current of known in- 

 tensity was allowed to flow through the test piece for a time correspond- 

 ing to the time occupied in a gradual load test, and the deflections 

 obtained were compared with those from a similar experiment when 

 a heavy current was used for a short interval of time. The heating 

 effect varies as I^Et, where I is the current, R the resistance and t the 

 time, and the experiments were adjusted so that the same heating effect 

 was produced as nearly as possible whatever the time might be. A 

 number of experiments were made, of which the following set may be 

 taken as a type: — 



TABLE VII. 



The times were taken by a stop-watch, and the values of k were 

 found in the usual manner, and due correction is made for the excess 

 of current in the first experiment. The maximum corrected value for 

 the first experiment is about three per cent greater than the coiTespond- 

 ing value for the second one, and this difference appears to be partly 

 due to the experimental difficulty of making and breaking the circuit 

 accurately in such a short period of time, when a stop-watch is used, 

 and also due to the fact that the correction factor D^ = -^o (1 + c^t) 



