137 



tricians never dealt with anything that 

 had constant coefficients of self-induction 

 or mutual induction, except measuring 

 instruments ; and in dealing with trans- 

 formers, incandescent lamps, and so on, 

 nothing of the sort came in. He thought 

 that all the comparisons in the paper 

 between direct and alternate-current 

 machines were based on an assumption 

 which made them misleading. The 

 assumption was that the direct machines 

 had the same number of armature turns 

 as the corresponding alternate machines. 

 In practice he did not think that any- 

 body ever made the air-gap otherwise 

 than equal to the breadth of the pole. 

 The output varied as the product of the 

 number of useful wires in a coil, and the 

 breadth of the pole-piece. The product 

 was greatest when they were equal, and, 

 consequently, designers always made 

 them equal. The fringe of useful field at 

 the edge of the pole-pieces must be 

 allowed for ; and in some cases the ques- 

 tion was complicated by the sides of the 

 coils being parallel, and the sides of the 



