138 ANNUAT, REPORT SMITHSONIAN INSTITUTION, 1911. 



II.—DUPLEX-DIPLEX TELEPHONY OVER WIRE CIRCUITS. 



Such has been the development of telephone engineering that at 

 present any proposal which requires for its success the supplanting of 

 the present low-frequency battery system would be most radical. It 

 would surely be admitted that any plan which permits the present 

 engineering telephone system to remain intact and superimpose 

 thereon additional telephone circuits would possess cardinal advan- 

 tages. Accordingly, the first preliminary experiments were directed 

 to the inquiry as to whether or not it is possible to superimpose upon 

 the minute telephonic currents now employed in telephony over wires, 

 electric waves of ultra-sound frequencies without causing prohibitive 

 interference with the battery telephone currents. Manifestly, this 

 fundamental point can best be determined by experiments, at the 

 generator itself, with the most sensitive part of the telephone equip- 

 ment, viz, the telephone receiver. Accordingly, experiments were 

 first conducted with various forms and types of telephone receivers in 

 connection with local circuits at the generator. Such is the sensibility 

 of the telephone receiver that it was thought possible that, although 

 currents of frequencies entirely above audition were applied to the 

 receiver from a dynamo as a source, there might be some frequency or 

 frequencies from the operation of the apparatus which would be within 

 the range of audition. Such was found, in fact, to be the case at cer- 

 tain critical frequencies of the machine, but they were of no practical 

 importance, as will be shown later. 



With a collection of telephone receivers ranging from about 50 to 

 over 8,000 ohms and of a variety of designs, a series of tests was made 

 under severe conditions to determine the above point. It was found, 

 in general, that alternating currents of frequencies ranging from 30,000 

 to 100,000 cycles per second, when coupled conductively, inductively, 

 or electrostatically to local circuits from the generator produced abso- 

 lutely no perceptible physiological effects in the receivers, excepting 

 only that at certain of the lower frequencies a distinct audible note 

 could be faintly heard in one of the receivers of about 250 ohms 

 resistance. 



A search for the cause of this note showed that it is due to a slight 

 variation of the amplitude of the high-frequency current of the gen- 

 erator, since no evidence of it could be detected on the battery tele- 

 phone side of the circuit. It appears to be caused by a very slight 

 vibration of the rotor as a whole in the magnetic field of the generator. 

 It was almost entirely removed by the simple device of opening out 

 the stators, winch increases the clearance and materially cuts down 

 the flux of the machine. In practice it is a distinct advantage, how- 

 ever, to have a trace of this note still left on the high-frequency side of 

 the circuit, otherwise there is no ready means of determining at the 



