Short Spark-ZHscharges. 6*97 



meter-telescope beside the microscope in such a position that 

 the two could be used as a binocular. 



In making a test the following method was used : — 

 («) Before each test unused portions of both discharge- 

 surfaces were brought together. The surfaces were polished 

 on a clean cloth buffing-wheel with soft dry rouge just before 

 closing the discharge-chamber. 



Fiff. 3. 



(b) The chamber was blown out with dust-free air from the 

 drving-tubes. 



The zoro was obtained by noting the position of the 

 Eringea in the microscope when upon separating P and N the 

 galvanometer showed that the testing circuit, shown on 

 the left of fig. 3, was broken. Throughout the test the zero 

 was re-determined each five or ten minutes. If the zero was 

 inconstant, or if a regular change large enough to necessitate 

 a correction was obtained, the testing was discontinued until 

 the zero was found to be sufficiently constant. 



A <<>n-iderable time was allowed to elapse between 

 changes of potential, and during this time the condenser was 

 recharged each four seconds. Thus during nine-tenths of the 

 rime it stood connected to tin- discharge-surfaces. 



