May 7, 1891] 



NATURE 



n 



I 



stop the wave, and it has long been known that, besides 

 the action of conductors as screens of electric force, dif- 

 ferent non-conductors act differently in reference to elec- 

 tric force by differing in specific inductive capacity. 

 Hence we might expect non-conductors to reflect these 

 waves, although the reflection would probably not be so 

 intense from them as from conductors. Hence this 

 question of how to reflect the waves is pretty easily 

 solved. We are acting still on the supposition that there 

 are waves at all. If electric force exist everywhere 

 simultaneously, of course there will be no waves to re- 

 flect, and, consequently, no loops and nodes produced by 

 the interference of the incident and reflected waves. 



(3) The third problem is : How can we expect to detect 

 where there are loops and where there are nodes ? Recall 

 the effects of electric force. It tends to move electrified 

 bodies. If, then, an electrified body were placed in a 

 loop, it would tend to vibrate up and down. This method 

 may possibly be employed at some future time, and it 

 may be part of the cause of photographic actions, for 

 these have recently been conclusively proved to be due to 

 electric force ; but the alternations of electric force from 

 positive to negative that have to be employed are so rapid 

 that no body large enough to be easily visible and electri- 

 fied to a reasonable extent could be expected to move 

 sufficiently to be visibly disturbed. It is possible that we 

 may find some way of detecting the vibrations hereby 

 given to the electrified ions in an electrolyte ; and it has 

 recently been stated that waves originated electrically 

 shake the elements in sensitive photographic films 

 sufficiently to cause changes that can be developed. The 

 other action of electric force is to produce an electric 

 current in a conductor and a resultant electrification of 

 the conductor. Two effects due to this action have 

 actually been used to detect the existence of the wave of 

 electric force sent out by a body alternately electrified 

 positively and negatively. One of these is the heating of 

 the conductor by the current. Several experimenters 

 have directly or indirectly used this way of detecting the 

 electric force. The other way, which has proved so far 

 the most sensitive of all, has been to use the electrifica- 

 tion of the conductor to cause a spark across an air- 

 space. This is the method Hertz originally employed. 

 A priori, one would not have expected it to be a delicate 

 method at all. It takes very considerable electric forces 

 to produce visible sparks. On the other hand, the time 

 the force need last in order to produce a spark is some- 

 thing very small indeed, and hitherto it has not been 

 possible to keep up the alternate electrifications for more 

 than a minute fraction of a second, and this is the reason 

 why other apparently more promising methods have 

 failed to be as sensitive as the method of producing 

 sparks. If two conductors be placed very close to one 

 another in such a direction that the electric force 

 is in the line joining them, their near surfaces will be 

 oppositely electrified when the electric force acts on 

 them, and we may expect that, if the force be great 

 enough, and the surfaces near enough, an electric spark 

 will pass from one to the other. This is roughly the 

 arrangement used by Hertz to detect whether there are 

 loops and nodes between the originator of the waves and 

 the reflector. 



Now arises the problem of how to electrify the body 

 alternately positively and negatively with sufficient 

 rapidity. How rapid is "with sufficient rapidity''.'' 

 To answer this we must form some estimate of how 

 rapidly we may expect the waves to be propagated. 

 According to Maxwell's theory, they should go at the 

 same rate as light, some 300 million of metres per second, 

 and it is evident that if we are going to test Maxwell's 

 theory we must make provision for sufficiently rapid 

 electric vibrations to give some result if the waves are 

 propagated at this enormous rate. The distance from a 

 node to a node is half the distance a wave travels during 



NO. 1 123, VOL. 44] 



a vibration. If we can produce vibrations at the rate of 

 300 million per second, a wave would go i metre during 

 a vibration, so that, with this enormous rate of alternation, 

 the distance from node to node would be 50 cm. We 

 might expect to be able to work on this scale very well, 

 or even on ten times this scale, i.e. with alternations at 

 the rate of 30 million per second, and 5 metres from 

 node to node, but hardly on a much larger scale than 

 this. It almost takes one's breath away to contemplate 

 the production of vibrations of this enormous rapidity. 

 Of course they are very much slower than those of light : 

 these latter are more than a million times as rapid ; but 

 300 million per second is enormously more rapid than 

 any audible sound, about a thousand times as fast as the 

 highest audible note. A short bar of metal vibrates 

 longitudinally very fast, but it would have to be about the 

 thousandth of a centimetre long, in order to vibrate 

 at the required rate. It would be almost hopeless by 

 mechanical means to produce electric alternations of this 

 frequency. Fortunately there is an electric method of 

 producing very rapid alternate electrifications. When a 

 Leyden jar is discharged through a wire of small resist- 

 ance, the self-induction of the current in this wire keeps 

 the current running after the jar is discharged, and re- 

 charges it in the opposite direction, to immediately 

 discharge back again, and so on through a series of 

 alternations. This action is quite intelligible on the 

 hypothesis that electrification consists in a strained 

 condition of the ether, which relieves itself by means of 

 the conductor. Just as a bent spring or other strained 

 body, when allowed suddenly to relieve itself, relieves 

 itself in a series of vibrations that gradually subside, 

 similarly the strain of the ether relieves itself in a series of 

 gradually subsiding vibrations. If the spring while relieving 

 itself has toovercomefrictional resistance,itsvibrationswiIl 

 rapidly subside ; and if the friction be sufficiently great, it 

 will not vibrate at all, but will gradually subside into its 

 position of equilibrium. In the same manner, if the re- 

 sistance to the relief of the strain of the medium, which 

 is off"ered by the conducting wire, re great, the vibrations 

 will subside rapidly, and if the resistance of the wire be 

 too great, there will not be any vibrations at all. Of 

 course, quite independently of all frictional and viscous 

 resistances, a vibrating spring, such as a tuning-fork that 

 is producing sound-waves in the air which carry the 

 energy of the fork away from it into the surrounding 

 medium, will gradually vibrate less and less. In the 

 same way, quite independently of the resistance of the 

 conducting wire, we must expect that, if a discharging 

 conductor produces electric waves, its vibrations must 

 gradually subside owing to its energy being gradually 

 transferred to the surrounding medium. As a conse- 

 quence of this the time that a Leyden jar takes to dis- 

 charge itself in this way may be very short indeed. It 

 may perform a good many oscillations in this very short 

 time, but then each oscillation takes a very very short 

 time. To get some idea of what quantities we are deal- 

 ing with, consider the rates of oscillation which would 

 give wave-lengths that were short enough to be con- 

 veniently dealt with in laboratories. 300 million per 

 second would give us waves one metre long ; consider 

 what is meant by 100 million per second. We may get 

 some conception of it by calculating the time correspond- 

 ing to 100 million seconds. It is more than 3 years and 

 2 months. The pendulum of a clock would have to 

 oscillate 3 years and 2 months before it would have per- 

 formed as many oscillations as we require to be per- 

 formed in one second. The pendulum of a clock left to 

 itself without weights or springs to drive it, and only 

 given a single impulse, would practically cease to vibrate 

 after it had performed 40 or 50 vibrations, unless it were 

 very heavy, i.e. had a great store of energy or were very 

 delicately suspended, and exposed only a small resistance 

 to the air. A light pendulum would be stopped by com- 



