140 THE MAGNETIC CIRCUIT IART. 46 



to those in the actual conductors a and 6 respectively. The addi- 

 tion of these fictitious conductors does not modify the armature 

 m.m.f. because they neutralize each other. The conductor of 

 may be considered as forming a turn with a, while b' forms a turn 

 with b. It will be seen that the m.m.f. of the coil aa' assists that 

 of the field coil A, while the m.m.f. of the coil W is opposite to that 

 of the field coil B. 



The armature current in the coil ab not only distorts the no- 

 load field, but also reduces the total flux per pole. This may be 

 seen by considering the flux hi the four parts of the air-gap, marked 

 x, y, x', and y', where x = x f and y=y'. The sum of the fluxes in 

 the portions y and y' is the same as without the armature current, 

 because the flux density hi the part y is increased by the same 

 amount by which it is reduced in the part y' (neglecting satura- 

 tion) . But in the parts x and x' the flux is reduced by the arma- 

 ture m.m.f., so that the total result over the pole-pitch is a reduc- 

 tion in the value of the no-load flux. The position of the armature 

 conductors and the direction of the armature currents have been 

 selected arbitrarily. They can be chosen so that the flux will be 

 crowded toward the left-hand tips of the poles, or so that the total 

 flux will be increased by the armature m.m.f., instead of being 

 reduced. The influence of the armature currents, in modifying the 

 value of the field flux and distorting it, is called the armature reac- 

 tion. The armature reaction is measured in ampere-turns, since 

 it is a magnetomotive force. 



In addition to the general distortion of the field by the arma- 

 ture currents, there is a local distortion around each armature 

 conductor. This distortion does not extend into the pole shoes, 

 but is limited to the slots and the air-gap; it is indicated in Fig. 36 

 by ripples in the flux around a and b. These ripples may be 

 regarded as a result of the superposition upon the main flux of the 

 local fluxes $ a and 0& excited by the armature currents. While 

 these local fluxes, shown by the dotted lines, have no real existence, 

 except around the end connections of the armature conductors, 

 it is convenient to consider them separately. They are purely 

 alternating fluxes, in phase with the currents with which they are 

 linked, so that they induce in the armature windings alternating 

 e.m.fs. hi a lagging phase quadrature with the currents. 



The effect of these local fluxes upon the voltage of the machine 

 is represented by a certain armature reactance, because the effect is 



