72 



NA TURE 



February i8, 1892 



magnetic waves, Hertz found that the half wave-length 

 in air was about 4*5 metres, the corrected period of the 

 vibrator used being about 2 x lO"** seconds. This gives 

 a velocity of propagation of 900/(2 x lo"^) (or 4*5 x iqI") 

 cms. per second, exceeding the velocity of light by about 

 50 per cent. For the wave-length in wires, however, he 

 found a value which gives a velocity of propagation 

 nearly equal to the velocity of light, when the correction 

 of the period for error in capacity is taken into account. 

 (Later, M. Poincard gives the half wave-length in air for 

 these experiments as 48 metres.) 



Herr Lechner, experimenting at Vienna, has also found a 

 velocity of propagation in wires very approximately equal 

 to the velocity of light, whicH thus confirms Hertz's 

 result. On the other hand, MM. Sarasin and de la Rive, 

 experimenting at Geneva in 1890, found that the wave- 

 length observed depends very much on the dimensions 

 of the resonator. But using an exciter exactly similar to 

 that of Hertz, and of the same dimensions, and a reiona- 

 tor 75 cms. in diameter, and therefore nearly an exact 

 copy of that employed by Hertz, they found a half wave- 

 length of 3 metres, instead of 48 metres as found by 

 Hertz. Thus there is a discrepancy between the two 

 results which it is difficult to explain. Hertz himself 

 gives a possible explanation, in a letter to M. Poincarc 

 which is quoted in a note on some recent experiments 

 which is printed as an appendix. On account of its 

 interest, we take the liberty here of translating the extract 

 quoted. It is to be noted that what is called the wave- 

 length here is the distance from node to node, or half the 

 complete wave-length. 



"It is difficult for me to believe that I have been 

 misled in the second method into finding 4*8 metres 

 instead of 3 metres ; but since the result of Messrs. 

 Sarasin and de la Rive has every theoretical appear- 

 ance of truth, I have endeavoured to find out the cause 

 of the difference. Here are two ways of explaining 

 it. The waves were produced between two parallel 

 walls of a room, and I have taken account of the reflec- 

 tive action of only one of them. Let us suppose, to begin 

 with, that the length of the room is an exact multiple of the 

 wave-length, say three wave-lengths. We shall have two 

 well-marked nodes at the exact distance. If the lengtfi 

 of the room is four wave-lengths, we shall have three 

 well-marked nodes. But if the length of the room is 

 intermediate between these, and nearer the former than 

 the latter, we shall have two less distinct nodes at a dis- 

 tance apart greater than a wave-length. This explana- 

 tion would appear to me satisfactory, if the difference 

 were not too great. 



"The other way of explaining the difference is this. 

 My reflecting plate of zinc was fixed in a niche in the 

 wall, and it is possible that the projecting parts of the 

 wall may have had the effect of carrying off the nodes to 

 a greater distance from the wall, and thus of giving too 

 great an apparent length as measured. But it is also 

 true that the niche was from 5 to 6 metres in width, and 

 it does not seem to me very probable that it can have 

 had any great effect.^ 



" I therefore cannot tell precisely the cause of my 

 error ; but I believe there must be some way of explaining 

 it For a long time I have sought in vain to find a prob- 

 able cause for the difference of velocity in air and in 

 wires ; and I had myself found, before Messrs. Sarasin 

 and de la Rive, that there is no difference for short waves 

 of 30 cms. in length. The results of these gentlemen, 

 however, give the same velocity for long waves, and con- 

 tradict my experiments " 



Connected with this point M. Poincard has some in- 

 structive remarks on what Messrs. Sarasin and de la Rive 

 have observed and called multiple resonance, and which 



' For an interesting discussion of the effects of rellecting plates of 

 <iifferent dimensions, see a paper by Mr. Trouton in the Pliil. Mag., 

 July 1891. 



has also been observed by Fitzgerald and Trouton. 

 Their supposition is that the exciter gives rise neither 

 to a single vibration of distinct period, nor to a limited 

 number of distinct vibrations, but rather to such a 

 complex of vibrations as would give a wide band ot 

 continuous spectrum. Thus all vibrations, agreeing with 

 possible modes of vibration of the resonator would be 

 reinforced. That this explanation is not borne out by 

 the theory is true, but on account of the incompleteness 

 of the theory it is not possible to attach much weight to 

 this fact. It is hard to believe that the vibrations can be 

 perfectly simple. 



Poincard proposes, however, the following explanation. 

 For various reasons, he thinks the logarithmic decrement 

 of the vibrations of the exciter is probably much greater 

 than that of the resonator, and so the vibrations of the 

 exciter diminish in amplitude more quickly than those 

 which by any cause are set up in the resonator. Thus 

 the resonator, being started by the exciter, would continue 

 its vibrations after those of the exciter had become in- 

 sensible, but would then vibrate in its own proper period, 

 thus giving vibrations of longer period and of greater 

 wave-length than those which excited it. The wave- 

 length, being determined by interference, and used with 

 the too short period of the exciter, would of course give 

 too great a velocity of propagation. With this explanation 

 Hertz has expressed himself as practically in accord, and 

 so a possible way out of the difficulty seems opened up. 

 As Hertz remarks, the oscillations of the exciter, re- 

 presented graphically, do not give a curve of sines pure 

 and simple, but a curve of sines the amplitude of which 

 gradually diminishes. Such an oscillation will cause all 

 resonators receiving it to vibrate, but those in tune with 

 the exciter more violently than the others. A mathe- 

 matical investigation of the point is given by M. Poincard, 

 which explains the result, shown by experiment, that the 

 apparent spectrum found by Sarasin and de la Rive 

 seems more extended when wires are connected to the 

 vibrator, than when the propagation takes place freely 

 in air. 



Whether this explanation be satisfactory or not, there 

 can be no doubt, on the whole, that the electromagnetic 

 theory of light is substantially true. The theory is far 

 from complete, and there are many outstanding points 

 which require further theoretical and experimental eluci- 

 dation. Some of these are touched on by Poincard in 

 his discussion of the field produced by exciters of different 

 forms, and the theory of the resonator, but especially in 

 a valuable series of notes which he has added to his 

 lectures. These deal with special topics, which are there 

 treated with more detail than was possible in the body of 

 the work. Such, for example, are his notes on multiple 

 resonance, the calculation of the period, and the propaga- 

 tion of waves in sinuous wires. 



This article has run to too great a length, and must here 

 close. M. Poincard's work ought to be read by everyone 

 interested in Maxwell's great scientific generalization — the 

 greatest, perhaps, ever made by a natural philosopher 

 since the days of Newton — and in its remarkable experi- 

 mental verification by Hertz. There never was, perhaps, 

 a time of greater mathematical and physical activity than 

 the present, but withal it is marked by a care for the 

 scientific student which no previous age ever displayed. 

 It is no small encouragement to humbler scientific 

 workers when masters of analysis like M. Poincard take 

 the trouble to publish, in a connected form, their lectures 

 and researches on the current scientific questions of the 

 day. Besides earning the gratitude of those who are 

 thus admitted within the circle of their pupils, by imme- 

 diately communicating their discoveries and expositions 

 in this manner to the general scientific public, they 

 multiply many-fold the direct effect of their work on 

 scientific progress. 



A. Gray. 



NO. 



I 164, VOL. 45] 



