76 ELECTRIC LIGHTING. 



terrupted, and this led to their arrangement in a circle. 

 Therefore an iron ring was used, surrounded transversely by 

 a coil which completely covered it, and subjected to the 

 action of an inducing magnet, simply magnetic at first, but 

 afterwards electro-magnetic. According to the effects ex- 

 plained on page 54, continuous currents should be obtained 

 without a commutator, their direction depending on the 

 rotatory movement. But it was soon noticed that opposing 

 actions were produced, and this led to the division of the 

 coil surrounding the ring into different sections, so that these 

 might be combined like the cells of a battery, and that the 

 poles of a powerful inducing electro-magnet might be made 

 to act upon them. Under these conditions it was, in fact 

 possible to place on the connecting wires of the several 

 sections of the coil, derivation wires communicating with 

 plates arranged round a circular drum, on which pressed 

 two springs to collect the current, and these springs, being 

 in this way placed in the position of a derivation between 

 two equal generators joined by their similar poles, would 

 transmit the current into the circuit without the necessity of 

 any inversions, and therefore without the use of commu- 

 tators. This was a very happy idea, and in it is contained 

 the essence of Gramme's invention. We shall presently see 

 that it has been applied to nearly all the new dynamo-electric 

 machines with continuous currents, a fact which proves its 

 importance, 



The effects produced in the Gramme invention are, how- 

 ever, somewhat complex, and are connecteo .with several 

 causes, the principal of which, as we have seen, is the action 

 of the magnetic poles on the several spires of the induced 

 coil ; but this action is itself double, for it may result from 

 the passage of the coils before these poles, as well as from 

 an action analagous to a displacement of these along a ring 

 differently polarized in its different points. Let us first study 

 the former effect, referring to Fig. 20, which theoretically 

 represents the outline of a Gramme machine, and let us sup 



