520 



NATURE 



[June 24, 1920 



field aj^^d to it. Upon the same core of finely 

 laminatecrlron there is a winding to carry high- 

 frequency current and one for the microphone 

 current. The microphone current as it varies 

 takes the iron to different magnetic states, alters 

 the permeability accordingly, and therefore varies 

 the self-inductance of the high-frequency coils. 

 Many matters of detail have had to be worked 

 out in perfecting the apparatus ; an important one 



Fig. 3. 



may be explained by aid of Fig. 3. Here A, B 

 are the terminals of the high-frequency circuit, 

 and C, D those of the direct-current or microphone 

 circuit. The iron core seen in the figure is part 

 of a closed magnetic circuit ; it is in two portions, 

 each of which carries half of the high-frequency 

 winding ; the microphone winding encircles both 

 portions of the core. Since the high-frequency 

 windings are wound to exert opposite magneto- 

 motive forces on the halves of the core, they in- 

 duce negligible high-frequency electromotive 

 forces in the microphone coil. 



The connections of the apparatus to the 

 alternator, the antenna, and the microphone 

 are shown in Fig. 4, which also introduces 

 further details, namely, the four condensers. 

 For simplicity the windings are represented 

 without their cores, though it must be re- 

 membered that the operation of the device 

 depends entirely upon the magnetic properties of 

 NO. 2643, VOL. 105] 



iron. The condensers Cj and C2 have as their 

 chief function the prevention of the flow of un- 

 desired acoustic currents in the high-frequency 

 windings. The condenser C^ tends to annul some 

 of the non-varying inductance in the circuit com- 

 prising the magnetic controller and the alternator, 

 and the condenser C4 appears to have been intro- 

 duced for phase adjustment, and enhances the 

 sensitiveness of the whole arrangement to changes 

 of microphone current. The performance of the 

 device is excellent; it is stated that a variation 

 of 0-2 ampere in the direct current through the 

 microphone has been made to alter the power 

 given to the antenna from about 6 kw. to 43 kw. 



It was by aid of this device that Ministers 

 in Washington conversed with President 

 Wilson in mid-Atlantic, the voice currents 

 reaching the magnetic controller after passing 

 over land wires from Washington to the 

 New Brunswick wireless station. It will be 

 noticed that in this method of varying an 

 element of the complete antenna circuit the varia- 

 tion is not effected directly by the voice, and in 

 this respect the method is unlike the former two 

 methods. 



A related class of methods of modulating the 

 oscillations of the antenna is that in which a 

 voice-varied mutual inductance is employed to 

 transfer the high-frequency energy from the 

 source to the antenna. Perhaps the nearest 

 approach to such a method is that of Kiihn, of 

 the Telefunken Co. of Germany, though in his 

 method the self-inductance of the circuits is varied 

 also. The method has not worked out so success- 

 fully as that of Alexanderson, and need not be 

 described here. 



A third class of methods of modulation aims 

 at varying the activity of the source of high- 

 frequency current ; this is in contrast with the 

 preceding methods, in which the functioning of 

 the source is not directly controlled. In the 

 methods to be described the point of application 

 of the control is, so to speak, behind the source, 

 the antenna being supposed to be in front. For 

 example, the direct current that creates the mag- 

 netic field of a high-frequency alternator, or the 

 direct current or voltage supplied to one of the 

 circuits of a triode oscillator — that is, an oscil- 

 latory circuit sustained in oscillation by means 

 of a three-electrode vacuum valve — might be 

 varied by the voice, and the high-frequency output 

 to the antenna be varied accordingly. Many very 

 miscellaneous schemes have been described ; the 

 difficulty is to make a representative selection. 

 The triode oscillator especially lends itself to a 

 multitude of ingenious designs. 



In Fig. 5 the circuits of a simple form of oscil- 

 lator are sketched. The coil marked L is con- 

 nected at one end to the anode, at the other to 

 the grid, of a triode, the filament being connected 

 to a tapping in the coil. The inductance L of the 

 coil and the electrical capacity C of the condenser 

 constitute the circuit in which oscillations are to 

 be maintained. The action of the circuits may 

 be explained broadly thus : Suppose an oscillatory 



