168 Mr. Herschel on certain motions produced in 
if the whole surface had been covered with the liquid. The 
motions however are confined to such portions of the mer- 
cury as are actually covered, all the rest remaining quite 
still : the effects too are modified by capillary action. 
12. When the circuit is completed in a conducting liquid, 
in the manner described in the beginning of this paper, the 
action is most forcible in the direct line joining the poles ; its 
violence diminishing as we recede from this line, though it 
continues Sensible to a great distance either way : and the 
course pursued by electricity in its passage through conduct- 
ing media, and its law of distribution within it, may in some 
degree be traced, by placing globules of mercury in different 
parts of a liquid ; when it will be plainly seen, that it is by no 
means confined, or nearly so, to the straight line between the 
poles, or to the surface of the conducting medium, but im- 
mediately on quitting the wires diffuses itself through the 
whole liquid, its density being a maximum in the space di- 
rectly between them, and diminishing rapidly as we recede 
from their line of junction. ^ 
13. The mechanical action appears (c ceteris paribus) to be 
proportional to the absolute quantity of electricity which 
passes, dato tempore, through a filament of the liquid at the 
point where it is exerted. The magnetic effect is proportional 
{cceteris paribus) to the absolute quantity of electricity in motion 
present at once, (or at any indivisible instant of time) in a 
given portion of the conducting wire, or within the sphere of 
action of the needle, that is, to its density.* To establish or 
* In these expressions I have conceived electricity as being transmitted through 
conductors according to the laws of a gas of high, but variable elasticity through 
pipes more or lesy obstructed, a supposition which will represent many of the 
