270 TRANSACTIONS OP SECTION G. 



where tlie signal vanished. In consequence of the large disturbing noises on 

 aircraft, this minimum method was not considered to be accurate enough. It 

 is necessai-y to use a system where signals can be heard whilst the bearing is 

 taken. The system devised is to u.'=e two coils at right angles, which can be 

 rotated together on a vertical axis. One coil alone is used first, and the system 

 rotated to be somewhere near the maximum of this single coil. Then the 

 second coil is introduced and its effects added to or substracted from those of 

 the first coil. If the first coil, is correctly on its maximum, then the second coH 

 will be on its minimum, and thus there "is no change of intensity on adding or 

 substracting the effects of this coil by a reversing switch. 

 This system was- applied to aeroplanes in two ways : 



(a) The Wing Coil Sj/stem. — In this case the aerials are fixed rigidly to the 

 aeroplane which have to be rotated in order to determine bearings. This system 

 is particularly useful in flying towards an objective where there is a wireless 

 station. 



{b) Botafing Coil System. — In this case rotating coils were placed in the 

 fuselage and rotated independently of the machine. Considerable difficuHies had 

 to be overcome to bring the R.A.F. system to a stage of perfection : 



{a) The external noises on an aeroplane. 



{b) The disturbances due to the magneto which produce considerable noise 

 in the telephones. The cause of this disturbance was traced to the emission of 

 very short waves by the magneto, the wave length being of the order of from 

 5 to 30 metres. 



(e) When using the fuselage coil system it was found that corrections had to 

 be applied for the deviation produced by the metaJ -work of the aircraft. 



In addition to the preceding difficulties there is another trouble in the 

 variation of bearings produced by atmospheric influence. The extent of these 

 variations is not large, possibly never more than about 3 deg. when using waves 

 of 2,000 metres and "upwards. 



In spite of the preceding difficulties excellent results have already been 

 obtained. A large number of flights have been made, and it has been" found 

 that the mean error of bearing is If deg. when using beacon stations whose 

 distance varies from 20 to 500 mHes. 



2. The Three-electrode Thermiomc Valve as an Alternating Current 

 Generator. By Professor C. L. Fortescue.^ 



1. The paper refers mainly to the theory of the valve and circuits as worked 

 out in the course of the development of Naval Wireless Telegraphy. This 

 development was relatively slow at first, but after the success of the first valve 

 transmitting sets fitted in 1917 became very much more rapid. 



2. The action of the valve and circuit in generating an alternating current 

 is first explained by the aid of a mechanical' analogy. An arrangement of a 

 spring-controlled piston, a system of water connections, and a douhle-ported 

 water valve can be imagined which has properties closely analogous to the 

 capacity-inductance circuit used with the thermionic valve. 



3. An approximate theory of the conditions that have to be satisfied is worked 

 out on a power basis. The necessary conditions for the maximum power output 

 from the tube are explained by 'means of the contour characteristics. 



4. The paper is concluded by a short description of certain of the valve 

 transmitting sets actually in use in the Naval Service. 



3. A Method of using Two Triode Valves in Parallel for Generating 

 Oscillations. By Professor W. H. Eccles and P. W. Jord.\n.^ 



The method described consists in arranging two triode tubes so that they 

 act in turn symmetrically upon an inductance-capacity circuit, one of the tubes 



' See The Electrician, Sept. 19, 1019, p. 294 ; Engivrcrivg. Oct. 10, 1919, p. 491. 

 - See The Electrician, Sept. 19, 1919, p. 292; Ifadio Ife'rirw, Vol, i., p. 80. 



