PIERCE. — THE COOPER HEWITT MERCURY INTERRUPTER. 393 



throwing a switch the Hewitt interrupter may be put in the place of the 

 iridium spark-gup. 



The receiving circuit B is at a distance of twenty meters from the 

 transmitting circuit. The receiving circuit consists of a variable air- 

 condenser in series with the 



receiving instrument and a 

 closed rectangular loop (2 m. 

 X 1 m.) of wire in a plane 

 parallel to the sending circuit. 

 When the receiving circuit is 

 brought to approximate reso- 

 nance with the sending circuit, 

 large deflections of the instru- 

 ment are obtained. In taking Figure IV. 

 readings the discharge was 



kept up during the period of swing of the instrument, and throws were 

 read. 



The following sets of readings (Table I) are given to show the uni- 

 formity and effectiveness of the mercury interrupter in comparison with 

 the iridium spark-gap. The columns of readings were taken in the 

 order given in the table. A rheostat in the primary circuit of the trans- 

 former was adjusted so that the mercury interrupter functioned vigo- 

 rously ; the set of readings in the first column were taken, then without 

 changing the receiving circuit of the rheostat, the iridium sjDark was put 

 in the place of the mercury interrupter, and was adjusted in length to 

 give its best deflection. With this adjustment readings in the second 

 column were taken. The third column shows another set of readings 

 with the mercury interrupter. The fourth column is with the spark- 

 gap. The readings are in centimeters with a scale distance of GO cm. 



Whence it is seen that the deflections with the mercury interrupter 

 are about four times as large as the deflections that could be obtained 

 under similar conditions with the iridium spark-gap. Other comparisons 

 with various capacities and inductances in the circuits gave likewise sub- 

 stantially larger deflections with the mercury interrupter than with the 

 spark-gap. With this interrupter the spark at the secondary of a Tesla 

 coil was also four or five times as long as that obtainable with the 

 iridium spark-gap. The increased spark length with the Tesla coil did 

 not, however, indicate an increase of integral effect in the secondary, as 

 the sparks may have been fewer in number. Hence, the experiment 

 with induction between loops was made. The result shows that the 



