OP ARTS AND SCIENCES. 249 



lighter and closely wound coil of No. 30 (B. & S.) copper wire, with 

 a suspending wire of the same gauge and 40 centimeters in length. 

 The resistance of the instrument was 206 ohms, that of the suspended 

 coil being 166 ohms. In some of the experiments the metallic vane, 

 which in the instrument as described by Kohlrausch dips into a vessel 

 of dilute sulphuric acid, was removed, and a needle point dipping into 

 mercury substituted, the connection with the telephone being made 

 through this. But the greater freedom from disturbances when the 

 vane was used made it preferable. The rapid alternation of currents 

 prevented any difficulty from polarization, as was shown by the fact 

 that with a source of sound of constant intensity the same results were 

 obtained whether dilute sulphuric acid or mercury was employed. A 

 reading telescope and scale graduated to millimeters served to deter- 

 mine the deflections of the dynamometer coil, the telescope being at a 

 distance of 1.5 meters from the mirror. 



The dynamometer was usually placed in the secondary circuit of a 

 small induction coil, the resistance of which was about 800 ohms. 

 The microphone transmitters were placed in the primary circuit, with 

 two Grenet cells arranged in parallel circuit. When the magneto 

 transmitter was used it was placed directly in circuit with the electro- 

 dynamometer. 



The dynamometer was so placed that, when no current was flowing, 

 the axis of its suspended coil was at right angles to the magnetic me- 

 ridian. No special care was taken to secure this accurately, as the 

 action of the deviating couple due to the earth's magnetism was found 

 to be quite negligible in comparison with variations caused by unavoid- 

 able differences in the intensity of the sounds acting upon the trans- 

 mitter. Deflections of the suspended coil were read by the use of a 

 mirror, telescope, and scale, as the ordinary method of bringing the 

 scale-reading back to zero by moving the torsion head of the instru- 

 ment could not be used, since it was impossible with the voice to 

 sustain a note for a sufficient length of time. Tlie deflections were 

 therefore read by the observer, and the strength of the current subse- 

 quently determined from these by the following method. A Daniel! 

 cell was placed in circuit with the dynamometer, using a mercury con- 

 tact to avoid the polarization which would ensue with dilute sulphuric 

 acid, and by means of a rheostat interposed in the circuit the resist- 

 ance was varied until deflections were obtained corresponding with 

 those given by the telephone currents. The electi'oraotive force of 

 the cell divided by the total resistance in the circuit gave the current 

 corresponding to any given deflection. A galvanometer was also 



