I 2 



NA TURE 



[July 7, 1923 



to assist the motion of the magnet. By means of a 

 small mirror fixed to the magnet, and a lamp and scale, 

 the building up of the motion from a small initial 

 amplitude is easily seen. 



With two such pendulums the accordant state can 

 be studied by eye observation. Dr. Winifred Leyshon 

 is engaged upon this task. As arranged for the investi- 

 gation one of the pendulums is made the master by 

 sending some of its current through an auxiliary winding 

 influencing the magnet of the other pendulum. The 

 frequency of either the master or of the servant pendu- 

 lum can be varied by the aid of a movable permanent 

 bar magnet placed near the oscillating magnet. Then 

 it is seen that as one natural period becomes nearly the 



Fig. 4. — Triode-sustained tuning-fork. 



same as the other the master catches hold of the 

 servant, compels it to abandon its own natural 

 period and to move in time with the master's — though 

 not necessarily in step. The amount by which the 

 servant is out of step depends upon the difference of 

 the natural periods and therefore can be regulated. 



These slow vibrations are seen and not heard ; but it 

 is also possible to use vibrators of acoustic frequency 

 and so make the according process evident to the ear. 

 A tuning-fork sustained by a triode is very effective as 

 the master oscillator. The circuit is shown in Fig. 4, 

 from which it will be seen that when the fork is vibrating 

 the induced electromotive force acting upon the grid 

 controls the anode current so as to sustain the motion. 



Fig. 5. — Triode electrical oscillator. 



(See Eccles and Jordan, " Sustaining the Vibration of 

 a Tuning-fork by a Triode Valve," The Electrician, 

 June 20, 1919.) 



On the other hand, an electrical oscillation, which is 

 independent of moving matter, makes a good servant 

 oscillator. Its circuit is shown in Fig. 5. The linkage 

 between the two oscillators is effected by passing some 

 of the current from the fork coils through an auxiliary 

 winding on the electrical oscillator. The fork is audible 

 when oscillating because it agitates the air; the elec- 

 trical oscillations can be made audible by inducing 

 currents in another circuit containing a loud-speaking 

 telephone, and their frequency can easily be altered 

 through a semitone or more by varying slightly the 

 capacity of the condenser shown in Fig. 5. Now, as 



NO. 2801, VOL. 112] 



the natural frequency of the electrical oscillator is made 

 to approach that of the fork, loud throbbings (called 

 " beats ") are heard, which become gradually slower 

 until at a certain point the master suddenly drags the 

 servant into time and the throbbings cease. If the 

 movement of the condenser is continued the natural 

 period of the electric oscillator is carried through 

 resonance and then beyond, and finally the servant 

 breaks away from the master and the throbbings indi- 

 cating their difference of frequency begin anew. 



This experiment is reminiscent of that of the two 

 air - blown organ pipes discussed by the late Lord 

 Rayleigh many years ago {Phil. Mag., 1879, Collected 

 Papers, vol. i. p. 409). Rayleigh showed that two 

 organ-pipes nearly in unison dragged each other into a 

 common frequency if brought into propinquity. 



The preceding experiments have carried us from 

 vibrations at 2 per second to vibrations at 200 per 

 second ; we now pass to the problem of accordance 

 when the vibrations are of frequency 200,000 per 

 second, such as are commonly used in wireless tele- 

 graphy and telephony. Such high frequencies are 

 neither seen nor heard, but can be detected by special 

 methods. The electrical oscillator used comprises 

 a triode and an inductance and capacity connected 

 as in Fig. 5 and chosen of suitable magnitudes. 

 The detecting apparatus is an inductance coil and 

 variable condenser connected to a crystal detector just 

 as in many a household crystal apparatus used for 

 listening to the broadcasting stations. A galvano- 

 meter is connected to the crystal and a spot of light 

 moves on the screen when the condenser is varied while 

 the triode apparatus is in action. A maximum deflexion 

 is soon found and then the receiver is in tune with 

 the triode oscillator. Another triode oscillator is now 

 substituted for the first and varied in frequency until 

 in tune with the crystal receiver. Clearly both triode 

 oscillators are now of approximately the same frequency. 

 Let them both be put into action simultaneously so as 

 to act upon the crystal circuit, and let a pair of auxiliary 

 coils, connected in series, be placed confronting the 

 respective triode oscillators in order to establish a 

 linkage. The crystal circuit is receiving energy from 

 both of the triode oscillators and actuates the galvano- 

 meter. The accordant state is then easily found by 

 varying one of the oscillators very slowly and watching 

 the spot of light. At the moment when the two oscil- 

 lators come within a certain frequency difference, they 

 suddenly pull into time and the spot of light gives a 

 sudden kick. This phenomena was discovered by 

 Dr. J. H. Vincent and described in the Physical Society 

 Proceedings (p. 84, Feb. 1920). One of his curves is 

 reproduced in Fig. 6. 



This curve illustrates that as the condenser of one 

 triode oscillator is increased the galvanometer in the 

 crystal circuit shows first an increase and then a very 

 sudden decrease of deflexion. The nearly vertical 

 parts of the curve are due to the establishment of 

 accordance. In a rough way one may explain the 

 phenomenon by saying that at the lowest point of the 

 curve, where there is a sharp cusp, the two oscillators 

 though vibrating in time with each other are oscillating 

 oppositely. In fact one oscillator is moving like the 

 front legs and the other like the hind legs of the dog 

 cited already. The curve or the experiment shows that 



