64 



Mr Palmer, Some Methods of Increasing 



terminal B. In the latter at G there was a small hole. B was 

 connected to the nearest metal of the terminal by means of a 

 wire W. The air-blast was formed by a pair of foot bellows which 

 were connected by a rubber tubing to the glass tube at R. 



The jars were next removed from the machine. When the 

 machine was running and the knobs about 6 cms. apart no 

 continuous brush discharge took place across the gap. On 

 bringing the end of a glass rod near the positive terminal and in 

 the gap, a large positive brush was formed extending about 

 three-quarters the distance across ; the negative brush was also 

 apparent and there was a narrow dark space between them. 

 Beyond a slight crackling no discharge took place when the 

 rod was brought near the negative terminal. 



(5) Suddenly altering the Shape of the Terminal. 



All the preceding methods of increasing the gap seemed to 

 depend in some way on a brush discharge being produced at 

 the same instant or just previously to the spark ; and since a 

 brush is easily produced on a point, it would follow that if one 

 were suddenly made on a round conductor the brush thereby 

 caused should force discharge. It was possible to lengthen the 

 spark by this means in the following way. A hollow ball of 

 indiarubber A was fastened in a cup of brass by means of sealing- 

 wax (Fig. 3) ; a hollow brass tube G was soldered to this cup and 



O 



To Jar 



Fig. 3. 



To Jar 



a hole was made through the sealing-wax and rubber by passing 

 a hot wire along C. If a glass rod B with end first pointed and 

 then slightly rounded in the Bunsen flame was pushed along C 

 into the ball, then at the point where it touched the rubber 

 a slight protuberance was made. A difficulty was experienced in 

 covering the ball with some conducting material which would 

 yield sufficiently when the ball was stretched. Tinfoil and Dutch- 

 metal were first tried but I was unable to place them on the ball 

 so that they altered with the shape of the rubber ; next a paste 

 of plumbago and gum was used and although imperfect the 

 phenomena were easy enough to obtain with it. The ball 

 (diameter 3'5 cms.) and tube formed one terminal of the Wims- 

 hurst, and the other pole was of brass and 2*9 cms. diameter. 



