Intelligence and Miscellaneous Articles, 215 



along their whole length with soft iron. The result was, that the 

 tabes under such circumstances scarcely obeyed the directing force 

 of the earth. This last experiment demonstrates, perhaps most 

 strikingly, how prejudicial to permanent steel magnets is the inte- 

 rior mass. 



Postscript. — Six months having elapsed since the above-mentioned 

 comparison (which took place shortly after the magnetizing of the 

 rods and tubes in question), and the above communication having 

 already been delivered to the editors of the Annalen, it recently 

 struck me to compare the same magnets again, as it certainly was 

 probable that hollow magnets would retain their magnetism better 

 than solid ones. The result was such as to far exceed my expecta- 

 tions ; for I found that now, with those larger magnets, the magne- 

 tism of the solid was to that of the hollow magnet as 1 : 2-5, and 

 the magnetism of the solid smaller magnet was to that of the hol- 

 low one in the ratio of 1 : 2-9. 



As, after this, more attention will probably be bestowed in future 

 on the manufacture of hollow permanent magnets for practical use, 

 I have resolved to devote to this subject a longer series of experi- 

 ments, and will in due time beg leave to communicate the results,— 

 Wiedemann's Annalen, 1880, No. 8, vol.x. pp. 694-696. 



ON AN ELECTRODYNAMICAL PARADOX. 

 BY M. GERARD-LESCUYER. 



Dynamoelectrical machines, of which the Gramme machine is 

 the best-known type, are reversible ; that is to say, if a current be 

 passed through them, they give motion, and may serve for the 

 transmission of force. Under the same conditions the same pro- 

 perties are possessed by magnetoelectrical machines with continuous 

 currents. In this there is nothing new. But if the current 

 produced by a dynamoelectrical be sent into a magnetoelectrical 

 machine, a strange phenomenon is witnessed, which we will 

 describe. 



As soon as the circuit is closed the magnetoelectrical machine 

 begins to move ; it tends to take a regulated velocity in accordance 

 with the intensity of the current by which it is excited ; but sud- 

 denly it slackens its speed, stops, and starts again in the opposite 

 direction, to stop again and rotate in the same direction as before. 

 In a word, it receives a regular reciprocating motion which lasts as 

 long as the current that produces it. 



What is the cause of this phenomenon ? 



Evidently the motive current must change in direction ; this is 

 proved by introducing a galvanometer into the circuit. But how 

 can this reversal of the current be produced while the velocity of 

 the generating machine (steam-eDgine, water- w T heel, &c.) does not 

 vary ? 



Some extraneous cause, then, must arise to reverse the polarities 

 of the inductors of the generating dynamoelectrical machine, so 

 that this machine may immediately give rise to a current of oppo- 

 site direction, which reverses the direction of rotation of the 



