84 Prof. Carl Barus on 



anemometer, least of all when its own song is drowned for 

 the ear by the noises round about. I therefore proposed to 

 myself in the first place to devise some means by which this 

 air-tone could be conveyed to any distance without change of 

 quality, and thereafter to endeavour to register this indication 

 automatically. It is only with the first of these problems 

 that the present paper is concerned. All attempts made at 

 the second, which is seriously difficult, have thus far been 

 failures*. 



The first condition appeared easy of accomplishment. It 

 seemed merely to be necessary to attach the wire to the plate 

 of a telephone in order to catch and transmit the sounds from 

 the roof to the office. However, telephonic devices over 

 which I laboured a long time failed to the last to give me any 

 result whatever. 



I then turned to the microphone, at first equally without 

 success, though finally an arrangement was developed which 

 behaved satisfactorily. I may in passing refer to certain of 

 the earlier forms of microphonic contact, some of which were 

 at times useful. The difficulty encountered arises from the 

 fact that a delicate microphonic contact is to be maintained 

 without breakage, at the end of a rapidly revolving and 

 necessarily vibrating whirling arm. Most devices give sounds 

 interrupted by a terrific racket at the receiver, which is very 

 trying to the ears. Apart from this the effect of centrifugal 

 force in varying the contact is to be counteracted, and the 

 air-resistance or the noises made by moving parts of the 

 microphone eliminated. 



In figure 1 aa is the revolving wire, attached at its upper 

 end to one arm of the whirling machine (not shown) , and at 

 the lower end to the metallic plate or sounding-board b, at 

 the end of the lower arm of the machine. Near the bottom of 

 the wire a very delicate brass spring cd is soldered on at c, 

 carrying a short platinum stylus at d. A similar spring, ef, 

 attached at/ to the sounding-board, carries at e a light pellicle 

 of carbon. The latter has been electroplated on one side 

 with copper, so that it can be soldered. The two springs are 

 so adjusted that the microphonic contact is just made at d, e, 

 and kept intact when e is slightly moved up and down, d 

 sliding on the flat face of e in this case. Moreover, as e is 

 relatively heavier than d, centrifugal force will tend to 

 straighten the spring/*? more than dc, and the contact is not 



* Unfortunately the ingenious devices of Prof. C. R. Cross and his 

 pupils (Troc. Am. Acad. xxv. p. 233 (1890); May 1892; Jan. 1893) 

 are not available here. 



