226 PEOCEEDINGS OF THE AMERICAN ACADEMY. 



observations given below were made with the platinum-faced balls. 

 For polishing the balls a chamois skin with a very small amount of 

 rouge was used, and sometimes they were finally cleansed with alcohol, 

 but this seemed to be no great advantage. 



On each side of the spark gap was a hard rubber vertical rod, 

 through which the conducting wire was jjassed and clamped by a 

 screw. These made a firm support, and held the balls rigorously at 

 the fixed distance apart. The leading wires from the induction coil 

 were soldered to closely fitting short brass tubes that were passed 

 over the conducting wires and rested against the rubber supports. 

 They always remained in the same position, and did not need to be 

 disturbed when the balls were removed for polishing. 



A large induction coil (53 cm. long and 19 cm. in diameter) was 

 used to charge the plates of the vibrator. To excite the coil five 

 storage cells were employed, which woi'ked with uniformly good re- 

 sults. The coil was capable of giving a spark 15 cm. long with this 

 source of electromotive force. A sparking distance of 4-6 ram. 

 was found most effective in producing oscillations in the secondary 

 circuit. 



The particular feature of the apparatus that applied directly to 

 the investigation was the secondary circuit. In the previous deter- 

 minations of the wave length due to the Hertzian vibration, the 

 arrangement originated by Hertz and modified by Lecher * and by 

 Sarsiu and De la Rive f has been generally employed. In this 

 arrangement secondary disks were placed face to face with the plates 

 of the vibrator, and to each secondary disk a long wire was attached, 

 and these wires then carried through the air parallel to each other, 

 with sometimes an additional disk on the free ends. 



With such an arrangement no exact adjustment of the length of 

 the secondary circuit was required in order to excite vigorous oscilla- 

 tion in it, for the direct electrostatic induction between the plates of 

 the primary and the disks on the ends of the secondary wires was so 

 great that powerful oscillations were produced along the secondary 

 wires, whatever their length might be, and several systems of waves 

 could be detected which seemed to give experimental grounds for 

 believing that the wave system sent out from the Hertzian vibrator 

 was very complex. 



The capacity of the vibrator is increased by the presence of these 



* Poggendorff, Annalen, XLI. 850 



t Archives des Sciences Physiques, XXIII. 113. 1890. 



