('If. VI EFFICIENCY 129 



The friction is made up of the friction of the brushes, 

 of the bearings, and of the air. The second of these- is the 

 most important, and if there is no side thrust due to a belt 

 or uneven pull of the magnets, will depend upon the 

 weight of the armature. 



It may not at first sight be obvious how the existence of 

 eddy currents produces a torque tending to oppose the 

 motion. But a little consideration will show that currents 

 induced in any part of the armature will have the same 

 effect as currents induced in the surface conductors, and 

 will tend to stop the motion. We may imagine all the 

 eddy currents to be concentrated in an imaginary con- 

 ductor short circuited on itself and not connected to the 

 surface conductors proper. The current in this conductor 

 is wasted, but it will exert a torque tending to stop the 

 motion, when there may be no current passing in the sur- 

 face conductors. If in is the induction factor of the waste 

 circuit, and r its resistance, the torque is given by 



W) 7? 



= 1'41 where n is the number of revolutions per 

 r 



second. Hence the eddy current torque varies as the 

 speed, and therefore the watt loss varies as the square of 

 the speed, m being constant. 



A convenient way of separating the different torque 

 losses is to run the machine by a second dynamo acting 

 as a motor. Observe first the current required to run the 

 motor alone, then couple on the dynamo to be tested and 

 run it without excitation. The increased current in the 

 motor will give the friction in the tested dynamo. Now 

 excite the magnets ; the increased motor current will then 

 enable us to find the increased torque required to turn the 



K 



