i20 Mr. G. V. MacLean on the 



kind and the same amount of capacity and self-induction 

 were used as in the coherer circuit. The coherer and 

 oscillator circuits were thus tuned to one another. 



The primary spark-gap of the Ruhmkorff coil was removed 

 to a distant mercury-break. The mercury-break was of 

 special construction. Three storage-cells drove a motor, 

 which in turn caused a plunger to play in and out of a 

 glass mercury-cup. On opposite sides of the glass cup were 

 secured a glass tube shoulder, 1 mm. or so above the level of 

 the mercury. To these glass shoulders were attached rubber 

 tubes, one leading from a water-tank and the other to a sink. 

 Water was syphoned from the tank, through the glass cup, 

 over the surface of the mercury- Thus the surface of the 

 mercury was always clean. A pinch- cock was fixed to the 

 first rubber-tube, so that the strength of the stream of the 

 running water could be regulated and kept constant. When 

 the stream of water was properly regulated, sparking at the 

 oscillator could be continued for hours at a time, the sparks 

 always being perfectly regular and unifoim. Jiefoie the 

 addition of the stream of water over the surface of the 

 mercury the break gave endless tiouble, needing attention 

 almost every two or three minutes : but under the conditions 

 adopted it did not require the slightest attention. However, 

 care had to be taken to prevent a too free flowing of the 

 water, for otherwise the nature of the sparking at the oscil- 

 lator was much altered, too much water acting like too little. 

 Wires ran from the break to the Iiuhmkoiff coil, and also to 

 a key at the coherer-carriage. Thus, from the coherer- 

 carriage one could control both the coherer and oscillator 

 circuits. 



The various parts of the oscillator circuit are shown in 

 Plate I. 



The motor and break were enclosed in a double box, the 

 space between the two boxes being packed with felt. This 

 was done in order to deaden the noise of the motor and 

 primary sparking. Such a proceeding is not essential to the 

 good working ot the coherer, but it enables one to detect at 

 once by the ear any variation in the secondary spaiking. ]t 

 thus leaves the eyes free to observe the movements of the 

 milliamperemeter needle. One soou becomes accustomed to 

 the characteristic crackling sound of the kind of sparks 

 required, and accordingly hardly ever needs to look at the 

 oscillator. 



The sparks produced at the oscillator must be continuous 

 and always the same, otherwise the waves set up will differ 

 from each other, and consequently the milliamperemeter- 





