MAGNETO-ELASTIC SOURCE OF NOISE 335 



seemed remarkable, and especially so in view of the fact that Bell, 

 the inventor of the telephone, in 1879 made passincj note ^ of an 

 experimental finding that probably was due in part to this effect. 

 Accordingly, the writers' interest was aroused to the extent that an 

 investigation was undertaken to learn the basic cause of the observed 

 result. 



In the light of subsequent knowledge it was surprising that some keen 

 observer had not predicted and demonstrated the effect as the natural 

 and necessary consequence of the researches of Ewing ^ and his pred- 

 ecessors upon the relation between state of magnetization and state 

 of stress of a ferromagnetic specimen. Apparently it remained for 

 von Hippel and Stierstadt ^ first to remark the phenomenon in 1931. 

 These men were unable to interpret the effect in simple terms, however, 

 and their reports presented a series of premature conclusions. \'on 

 Auwers * alone recognized the effect as capable of complete and 

 satisfying explanation on the basis of magneto-elasticity, but he chose 

 a method of establishing this, the interpretation of which was quite 

 involved. For their own satisfaction in comprehending the phenom- 

 enon the present writers were led to conduct a series of experiments 

 of qualitative character. It was felt that knowledge and clear under- 

 standing of the effect should be of immediate interest and value to 

 workers in the general field of electrical communication in the United 

 States. 



With the aid of an amplifier having a gain of 110 decibels and 

 terminating in a loud speaker it was practicable to conduct the experi- 

 mentation with specimens of table-top dimensions. This amplifier 

 had an input impedance very much higher than that of any of the 

 specimens, and the response of the speaker therefore was proportional 

 to the voltages generated by the specimens. A stretched iron wire 

 three feet long would yield a clear sound in the speaker when its ends 

 were connected to the input of the amplifier while it was being me- 

 chanically stimulated by plucking or bowing, and the sound from the 

 speaker would be closely of the same quality as that heard by direct 

 air transmission from the vibrating specimen. With this arrangement 

 it was possible to detect any change in the magnitude of the effect as 

 great as two to one simply by observing the loudness of the sound. 



It was verified immediately that the effect must be dependent upon 

 the property of ferro-magnetism. Taut wires of soft iron, tempered 

 steel, or nichrome ; rods of soft steel or permalloy— all produced strong 

 sounds in the speaker when their ends were connected to the input of 

 the amplifier while they were stimulated to vibration by bowing, 

 plucking, or tapping. With wires of copper or brass, or with a rod of 



