440 Prof. W. C. Ronteen on Electrical 



o 



mature reflection, resolved to select for my purpose a kind of 

 discharge which has hitherto been but little studied, viz. the 

 so-called discharge by convection, such as is known to take 

 place between a very sharp point and a large smooth plate. 

 I believe I may indeed attribute it to this selection, if I finally 

 succeeded in discovering the relation sought. 



The method of experiment at last found serviceable was the 

 following. A Holtz machine was kept in action with as con- 

 stant and high velocity of rotation of its disk as possible, by 

 means of a Schmidt water motor. One of the electrodes was 

 connected with the earth by the gas-pipes ; and from the other a 

 copper wire covered with gutta percha led to the inside coat- 

 ings of two Leyden jars constructed according to W. Thom- 

 son's plan, of well-insulating glass and containing sulphuric 

 acid, the outside coatings of which were led away to earth. 

 These jars formed an electrical reservoir of considerable 

 capacity, and were intended to diminish as much as pos- 

 sible the variations of potential which might be occasioned 

 by an irregular development of electricity by the machine. 

 Behind these jars the wire was divided. One branch went 

 to a narrow glass tube filled with glycerine, which served 

 as a rheostat ; by lowering into or drawing out of it a metallic 

 conductor to earth, the resistance of the glycerine could be 

 fixedly diminished or augmented. The other branch led first 

 to the point in the discharge-apparatus, and from that to an 

 electrometer constructed expressly for the investigation. 



The discharge-apparatus consisted of the following parts : — - 

 A vertical brass rod, provided beneath with a gilt sewing- 

 needle, passed well insulated through the tubular neck of a 

 glass bell, which was fitted air-tight upon the plate of an air- 

 pump. In the space enclosed by the bell stood, carefully 

 insulated from the plate, at a distance of 19*3 millims., with 

 its centre opposite the point, a polished brass disk of 132 mil- 

 lims. diameter; this was in conducting connexion with one 

 end of the coiled wire of an extremely delicate mirror-galva- 

 nometer consisting of a great number of turns ; the other ex- 

 tremity of the wire led to the gas-pipes. By an air-pump and 

 further suitable arrangements the bell could be filled with 

 different gases at various pressures determined by a mano- 

 meter. 



The electrometer made use of proved, it is true, serviceable 

 for the present investigation; but it still has some defects, 

 which must be removed ; I am therefore engaged in construct- 

 ing a better apparatus, of which I hope subsequently to give 

 an account. I will only mention further that it was arranged 

 after the manner of Thomson's quadrant-electrometer, and that 



