380 BELL SYSTEM TECHNICAL JOURNAL 



suitable positive potential by an external "B" battery, and (3) a grid elec- 

 trode placed close to the filament in the path of electron flow to the plate 

 electrode and used so as to exercise a very sensitive control of the electron 

 stream. This third element was deForest's pioneering contribution to the 

 art. In using the Audion as a telephone amplifier, the grid and filament 

 terminals serve as the input terminals and the plate and filament terminals 

 as the output terminals. Under proper operating conditions, variations in 

 input voltage applied to the grid circuit so afifect the flow of electrons as to 

 produce amplified voltages in the plate or output circuit. Using energy 

 drawn from the plate battery, an increase in energy is delivered to the out- 

 put line. In deForest's recent use of the Audion as a radio receiving am- 

 plifier the grid circuit included a series condenser; this was also included in 

 the arrangement offered as a telephone amplifier. 



The laboratory demonstration had been arranged with Mr. Carty by 

 John Stone Stone, a mutual friend and an independent research worker who 

 had acquired a theoretical knowledge of telephone transmission problems 

 when he was a member of the Boston headquarters staff of the telephone 

 company during the nineties. Colpitts and Richards participated in the 

 complete demonstration and Jewett in the final stages. This demonstration 

 of the Audion was entirely qualitative in scope, under simple but adequate 

 circuit conditions. With very low input levels, the speech currents were 

 greatly amplified without perceptible impairment in intelligibility. How- 

 ever, when the speech input approached the levels that would be encoun- 

 tered in any commercial use of repeaters, the amplification was greatly 

 reduced and very noticeable distortion and noise resulted. Under these 

 conditions, a blue haze was prone to appear in the tube and it disappeared 

 when the input level was reduced. As the plate potential was progressively 

 raised, a permanent condition of blue haze developed and the device ceased 

 to amplify. 



On the following day, November 1, the Audion was called to Arnold's 

 attention. He promptly repeated and extended the experiments, using 

 deForest's tubes and auxiliary apparatus which had been loaned for that 

 purpose. Arnold's broad training in electron physics enabled him without 

 study or delay to explain the blue haze phenomena and to prescribe a 

 remedy. The keynote of the explanation was that the blue haze was due 

 to ionization of gas present in the device and the remedy was to secure a 

 much higher vacuum. The medium vacuum in the Audion test samples 

 was a normal result of the best evacuation processes then used by incan- 

 descent lamp manufacturers. Better results could be secured by laboratory 

 processes known to research physicists and still better results could be ex- 

 pected from a new type of molecular pump then recently described in 

 technical literature. Arnold's preliminary analysis also indicated that a 



