lO 



NA TURE 



[May 3, 1S94 



Calculation shows us that the rate of spreading 

 is equal to the ratio of the units, that is to say, to the 

 velocity of light. 



These alternating currents are a kind of electrical i 

 vibrations ; but are these vibrations longitudinal like > 

 those of sound, or transversal like those of Fresnel's ' 

 " ether" ? In the case of sound the air undergoes conden- 

 sation and rarefaction, alternatively. On the contrary, 

 Fresnel's ether, when vibrating, behaves as if it were 

 formed of incompressible layers, capable only of sliding 

 one over the other. If there were open currents, the 

 electricity going from one extremity to the other of one 

 of these currents would accumulate at one of the 

 extremities ; it would condense or rarefy itself like air ; its 

 vibrations would be longitudinal. But Maxwell admits 

 only closed currents ; this accumulation is impossible, 

 and electricity behaves like Fresnel's incompressible 

 ether ; its vibrations are transversal. 



Experimental Verification. 



So we find again all the results of the undulatory 

 theory. But this was, however, not enough to induce 

 the physicists, who were more charmed than convinced, 

 to accept Maxwell's ideas. All that could be said in 

 their favour was that they did not contradict any of the 

 obser\ed facts, and that it was a great pity if they were 

 not true. But experimental confirmation was wanting ; 

 it had to be waited for during twenty-five years. 



A divergence had to be found between the old theory 

 and Maxwell's, which was not too delicate for our rough 

 means of investigation. There was only one which 

 afforded an expert me ntum crucis. 



The old electrodynamics required electromagnetic in- 

 duction to be produced instantaneously ; but according to 

 the new doctrine it must, on the contrary, be propagated 

 with the velocity of light. 



The question was therefore to measure, or at least to 

 ascertain, the rate of propagation of inductive effects ; 

 this has been done by the illustrious German physicist. 

 Hertz, by the method of interferences. 



This method is well known in its applications to 

 optical phenomena. Two luminous rays issuing from the 

 same source interfere when they meet at the same point 

 after having followed different paths. If the difference 

 of these paths is equal to the length of a wave — that is 

 to say, to the path traversed during one period, or a 

 whole number of wave-lengths— one of the vibrations is 

 later than another by a whole number of periods ; the 

 two vibrations are therefore at the same phase, they are 

 in the same direction, and they reinforce each other. 



If, on the contrary, the dilTerence of path of the two 

 rays is equal to an odd number of half wave-lengths, the 

 two vibrations are in contrary directions, and they 

 neutralise one another. 



The luminous waves arc not the only ones susceptible 

 to interference ; all periodic and alternating phenomena 

 propagated with a finite velocity will produce analogous 

 effects. It happens with sound. It ought to happen with 

 electrodynamic induction, if the velocity of propagation is 

 finite ; but if, on the contrary, the propagation be 

 instantaneous, there will not be any interference. 



But one cannot put these interferences to the proof 

 if the wave-length is greater than our laboratories, or 

 greater than the space that the induction can traverse 

 without becoming too feeble. Currents of very short 

 period are absolutely essential. 



Electric Exciters. 



Let us first see how they may be obtained with the 

 help of an apparatus which is a veritable electric 

 pendulum. .Suppose two conductors united by a wire; 

 if they arc not of the same potential, the electric equi- 

 librium is broken in the same way as the mechanical 

 equilibrium is deranged when a pendulum is swung from 



NO. 1279, VOL. 50] 



the vertical. In the one case as in the other, the equi- 

 librium tends to re-establish itself 



A current circulates in the wire, and tends to equalise 

 the potential of the two conductors in the same way as a 

 pendulum seeks the vertical. Hut the pendulum will 

 not stop in its position of equilibrium; having acquired 

 a certain velocity, it passes this position because of its 

 inertia. Similarly, when our conductors are discharged, 

 the electric equilibrium momentarily re-established, will 

 not maintain itself, and will be destroyed by a cause 

 analogous to inertia ; this cause is self-induction. We 

 know that when a current stops it gives rise in the 

 adjacent wires to an induced current in the same 

 direction. The same effect even is produced in the wire 

 in which the induction current circulates, which finds 

 itself, so to speak, continued by the induced current. 



In other words, a current will persist after the dis- 

 appearance of the cause which produced it, as a moving 

 body does not stop when the force, which had put it in 

 motion, ceases to act. 



When the two potentials shall have become equal, the 

 current will therefore continue in the same direction, and 

 will make the two conductors take opposite charges to 

 those which they had to start with. 



In this case, as in that of the pendulum, the place of 

 equilibrium is passed : in order to re-establish it, a back- 

 ward movement is necessary. 



When the equilibrium is regained, the same cause im- 

 mediately destroys it, and the oscillations continue with- 

 out ceasing. 



Calculation shows that the duration depends on the 

 capacity of the conductors ; it suffices, therefore, to 

 diminish sufficiently this capacity, which is easy, to have 

 an electric pendulum susceptible of producing alternating 

 currents of extreme rapidity. 



All this was well established by Lord Kelvin's theories 

 and by Feddersen's experiments on the oscillating 

 discharge of the Leyden jar. It is, therefore, not this 

 which constitutes the original idea of Hertz. 



But it is not sufficient to construct a pendulum: it 

 must also be put into movement. For this, it is neces- 

 sary for some agent to move it from its position of 

 equilibrium, and then to stop abruptly — I mean to s.ay, in 

 a time very short in relation to the duration of a period ; 

 otherwise the penduhmi will not oscillate. 



If, for example, we move a pendulum from its vertical 

 position with the hand, and then, instead of loosing it 

 suddenly, we let the arm relax slowly without unclaspinj; 

 the fingers, the pendulum, still supported, will arrive at 

 its place of equilibrium without velocity, and will not 

 pass it. 



We see then, that with periods of a hundred-millionth 

 of a second, no system of mechanical unclamping could 

 work, however rapid it might appear to us with regard to 

 our usual units of time. This is the way in which Hertz 

 has solved the problem. 



Taking again our electric pendulum, let us make in the 

 wire, which joins the two conductors, a cut of some milli- 

 metres. This cut divides our apparatus into two symmetric 

 halves, which we will put in communication with the two 

 poles of a Ruhmkorff coil. I'he induced current will 

 charge our two conductors, and the difference of their 

 potential will increase with a relative slowness. 



At first the cut will stop the conductors from dis- 

 charging themselves. The air plays the part of an 

 insulator, and keeps our pendulum away from its position 

 of equilibrium. 



But when the difference of potential becomes large 

 enough, the jar spark will pass, and will make a 

 way for the electricity accumulated on the conductors. 

 The cut will all at once cease to act as an insulator, and 

 by a sort of electric unclamping, our pendulum will 

 be freed from the cause which prevented it return- 

 ing to its equilibrium. If the complex conditions, well 



