DOUBLE-CUERENT MAGNETO-PAEADTC APPARATUS. 



279 



duced by movements of approximation or removal operating 

 mechanically."^ 



This understood, when we interrupt the current at the moment 

 of its maximum, it is as if we should interrupt a constant current, 

 and it produces an extra current, the curve of which lias been 

 already given, and which we here will reproduce (fig. 65). 



(^ 



X7 



Fig. 65. Fig. 66. 



Fig. 65. — Above : approximative representation of tlie current Induced by the movement of tlie reels of 

 a Clarlse's machine ; below : a mure exact representation. 



Fig. 66. — Above: representation of the extra current of interruption; below: representation of the 

 current induced by this extra current. 



" It would seem that the arcs, which intercept at the point B, 

 should be in excellent condition of form to give rise to intense 

 induced currents. That is true, but it must be remembered that 

 unity of time is here represented by a very short space, so that, if 

 we wish to represent the interval of time Aq Bq with the same unity 

 that has been employed, — for example, in fig. 65, to represent 

 the cessation of an interrupted current, — it would be necessary 

 to give to Aq Bq a length several hundred times greater. It is 

 manifest that, when thus extended, the curve of the current of the 

 reel would present only an insensible sinuosity, when compared 

 with that of the current of interruption. 



* It may be inquired how it happens 

 that an extra current is not produced at 

 the moment when a Clarke's reel, for 

 example, jjasses before the pole of a 

 magnet, since at that moment the current 

 suddenly changes its direction, as has 

 been stated above ; and it would seem 

 that a change of direction of this kind 

 should be even more efficacious than an 

 interruption. But, in fact, the sudden 

 change represented by the curve is only a 

 fiction ; in reality, the change is relntively 

 rapid but not sudden, because we have 

 not mathematical points which pass one 



before the other ; and the relatively con- 

 siderable dimensions of the reels of the 

 magnets canse the action to diminish 

 from a certain movement, and to increase 

 again. The decrease wliicla really occurs in 

 the vicinity of Bg ffig. 65), however rapid 

 it may be, is still very slow as compared 

 to the return of the natural state in a 

 circuit that is interrupted. 



Fig. 65 shows, by its lower curve, what 

 would be the actual distribution of the 

 intensity of the cm-rents in one of Clarke's 

 reels. 



