ELECTRICAL MEASUREMENTS. 265 



to the surface so that we have not to deal with volume in the 

 case of conducting bodies but with surface only. The super- 

 ficial density is the quantity of electricity per unit of surface. 

 Thus electrical Density is a subsidiary magnitude connected 

 with Quantity. Then in connection with Capacity there is a 

 magnitude which may be regarded as a subsidiary one. You 

 are well aware that the capacity of a given conductor, of 

 this cylinder for instance, that is the quantity of electricity 

 which it must take up in order to have a charge of given 

 potential, depends partly on the form and partly on the 

 dimensions, but also on the distance of surrounding bodies 

 and on the nature of the substance which separates the 

 conductor we are discussing from the surrounding . bodies. 

 There is then a property of the surrounding medium which 

 has something to do with determining the capacity of a given 

 conductor. This cylinder for instance is surrounded by 

 conductors such as the walls of the room, the ceiling, the 

 table, and so on, as well as by other nearer bodies, and it is 

 separated from these conductors by air. If we were to 

 replace the air by another insulating medium, the electrical 

 capacity of the cylinder would be changed ; and that 

 property of the surrounding medium which enables it to 

 produce an effect on the capacity of a given conductor is 

 called the specific inductive capacity of the medium or the 

 dielectric co-efficient, these being synonymous names for the 

 same property. 



The measurement of these different magnitudes constitutes 

 the problem that we have to discuss in talking of electro- 

 static measurements. In the case of any measurement we 

 may adopt what we may call a direct process, or an indirect 

 one, and to make more clear what I mean by this distinction 

 I may remind you of familiar processes in other cases. I 

 call a direct measurement one such as an ordinary measure- 

 ment of length. If we want to measure the length of this 

 table we take a foot-rule or any other standard we may agree 

 upon, and by measuring off this standard length over and over 

 again we estimate the length of the table in terms of the 

 adopted unit : but in the case of electricity such a process as 

 this is not available. One cannot take a standard quantity 

 of electricity and compare this by a direct process with any 

 other quantity that we want to measure. We have to adopt 

 a process less direct, in which we observe the effect which the 



