86 Prof. Carl Barus on 



explanation. After having made a delicate contact I noticed 

 a low sonorous sound in the telephone, which continued with 

 undiminished intensity for fully five minutes. No spring could 

 have vibrated for this length of time, so that the acoustic 

 apparatus must have contained its own motor. The sound 

 ceased only when contact was broken, but on succeeding days 

 I could not reproduce it. 



4. A second microphonic contact with which there is 

 usually less racket than in the preceding case is shown in 

 figure 2 in plan. The sounding wire is shown in cross 



Fig. 2.— Horizontal Spring Microphonic Contact on Air- Wire. 



f '/c c d \ 



section at a ; h and k are two very delicate flat springs 

 attached to the wire a at its lower end by two insulating plates 

 of hard rubber b, b', and the insulated clamp screws c, c\ The 

 free ends of these springs carry two small carbon pellicles, e 

 and /, centrally perforated so as to hold the shouldered rod 

 of graphite g loosely between them. The current of four 

 Leclanche cells passes through a } A, e, g, f, k, /3 (a and /3 being 

 the terminals) to the enclosed primary of the transformer, and 

 thence back to the battery. The rod g may also fit into cavities 

 in e and /, and be held in place by the very gentle clutch of 

 the springs h and k ; or g may be replaced by a short wire of 

 platinum. There is a groove in the middle of g to attach 

 small weights to vary the contact ; h and k are set radially, 

 so as not to be influenced by centrifugal force. 



No effect was obtained when either of the microphonic 

 contacts were placed on the sounding-board in which the 

 lower end of the air-wire terminates. It seems essential, 

 therefore, to tap the sound from the wire itself. A weighted 

 sounding-board gives no novel results. A gridiron of wires 

 produced air-tones which the ear appreciated with an in- 

 tensity proportional to the number of wires, but the effect 

 in the telephone did not exceed that of a single wire. Strips 

 of metal in place of the wires were also ineffectual, and it 

 was difficult to keep the sharp edge turned in the direction 

 of motion. In this case, too, the bulging of the central parts 

 of the revolving filament proved to be a serious annoyance. 



