196 Mr. J. J. Waterston on some Electrical 



point, the number in a square inch is oc -, and the force on a 



et 



square inch is as -$; but the number being cc -, the force of 

 ec at 



each one is a -• 

 a 



Thus it appears that the fundamental law of action is that the 

 static longitudinal pull at p is inversely as the area of the infini- 

 tesimal cylinder udr. Now to fix this in the mind, we may ima- 

 gine the cylinder to be a reality, and, however great or however 

 small, to contain the same constant quantity of an elastic fluid, 

 which will thus be more compressed when a is small than when 

 it is large. Further, let us suppose a valve or pressure-gauge 

 at p, or rather two small pistons, acting one in the direction pl v 

 and the other in the direction pl^, under the influence of the 

 elastic fluid contained in the cylinder. This elastic fluid, when 

 of double density, exerts double pressure, according to the law 

 of Mariotte ; so that it is not difficult, by keeping this little arti- 

 ficial arrangement in view, to retain a distinct idea of the distri- 

 bution of force in an electric line. 



How Nature accomplishes this curious arrangement of force, 

 and supports it from two excited material terminals, is of course 

 a profound mystery ; but that it does exist is, I submit, a fact 

 inductively established. We have no choice, we cannot deviate 

 from it, if we regard accuracy in the conception of the potential 

 nature of electricity. It is a theory, not an hypothesis; a 

 system proved from Harris's observations, not a system coined in 

 the imagination. 



56. With regard to practically determining the mechanical 

 equivalent of a given charge (that is, a given quantity of elec- 

 tricity on a given surface), the only difficulty is from the portion 

 that is free — how to ascertain its amount and allow for it. 



The following occurs as perhaps practicable with Harris's appa- 

 ratus. [The unit-jar as a measure of quantity seems to be unex- 

 ceptionable ; would it not be well to have an arrangement for an 

 absolute-xmit standard measure ?] 



Suppose we wish to compute the integral of a charge of g unit- 

 jars given to a sphere 10 inches in diameter, having a surface of 

 314 square inches. Let B be a circular conducting plate of 100 

 square inches mounted horizontally in a perfectly insulated con- 

 dition, and at the distance of an inch let A, another circular plate 

 of the same size, be suspended by newly high-dried silk lines to 

 one side of a balance. Let a charge be given to B by unit-jar, while 

 a free conductor touches A ; and after removing the free conductor 

 from A, let the weight required to raise A be measured as in 

 Harris's experiments. Next let B be touched with a free conduc- 



