136 AUSTIN: RADIOTELEGRAPHIC SIGNALS 



strength of signal or galvanometer deflection we can determine at 

 once the intensity of the electric field at the receiving point. 



Some detectors, for example the perikon, ^ can not be calibrated 

 with the microammeter as their adjustment is disturbed by the 

 amount of high frequency energy required to make a suitable 

 deflection. For such detectors the following method of calibra- 

 tion may be used. The middle circuit may be opened and leads 

 of about 50 cm. in length brought out to a coil of two or three 

 turns of wire wound on a hard rubber core. On this same core, 

 at a sufficient distance from the first coil to give proper coupling, 

 a second coil of say 0.1 m.h. is wound, and to this coil is connected 

 the detector to be tested with suitable stopping condenser and 

 telephones with an audibility meter of the type described in this 

 paper shunted across them. In this case the audibility meter 

 takes the place of the microammeter of the first method and the 

 coupling should be adjusted and made permanent for an audibility 

 of about 100 times with an average detector for a wattmeter 

 reading of 0.005 watt. 



(B) An audibility meter. For the determination of the loud- 

 ness of the signals in the telephone by the shunted telephone 

 method, the most convenient form of shunt resistance box is 

 one in which all the resistances are controlled by a single dial 

 arm. In general, thirty to forty studs are sufficient, giving differ- 

 ent degrees of audibility varying by approximately 20 per cent. 

 The exact inductive resistance of the telephone, of course, varies 

 with the spark frequency, and to a lesser extent with the capacity 

 used in the stopping condenser which seems to affect the wave 

 form of the rectified pulses. The inductive resistance of modern 

 2500 ohm telephones used with a stopping condenser of not more 

 than 0.02 microfarad at a thousand sparks per second, is approxi- 

 mately 5000 ohms. With the old fashioned low frequency spark, 

 the inductive resistance difl"ers very little from the direct current 

 resistance. If the resistance box is made for use with a definite 

 telephone and a definite spark frequency it may be very conven- 

 iently marked in audibility instead of in resistance. 



8 The silicon on the contrary is very stable. 



