1897.1 HOUSTON, KEXXELLY — THE PATH OF A CURREXT. 153 



short lengths of each of the two wires; namely, fg and hj, each one 

 metre long, are insulated from the rest of the system, and elec- 

 trically charged by a momentary connection to a dynamo or other 

 •suitable source, to a pressure of say, looo volts. An electric flux 

 will thus be established between the two short lengths of wire of the 

 type represented in Fig. 3, and shown in Fig. 4, by the arrows; fg, 

 being positive, and hj, negative. Strictly speaking, the flux dis- 

 turbance would not remain in parallel planes at the ends of the 

 short lengths, but would bulge outwards considerably, from the ends ; 

 but this peculiarity is of no consequence to what follows, and we 

 may, therefore, suppose that the simpler geometrical distribution is 

 preserved. If the metre lengths at E, be perfectly insulated from 

 -each other, the block of electric flux resident in the ether between 

 them would be indefinitely maintained at 1000 volts pressure. 



Suppose, however, that at some instant of time the discontinuities 

 existing between the metre lengths of the conductors and the rest of 

 the system are suddenly bridged over. In other words, the metre- 

 lengths are connected electrically at both ends to the rest of the 

 circuit. Then instantly the flux tends to rush towards the ends of 

 the circuit as represented by the arrows k and 1, in Fig. 5. The 

 metre block of flux instantly subdivides into two metre blocks, 

 each under 500 volts pressure, and each with half the original 

 density and, therefore, one-quarter of the original energy. At the 

 same time, the moment that the flux commences to run, a magnetic 

 flux distribution is brought into existence ; for, as we have already 

 mentioned, a motion of electric flux can never occur without pro- 

 ducing magnetic flux. While then the metre block of Fig. 5, 

 divides into two separate metre blocks, moving in opposite direc- 

 tions, as in Fig. 6, each block becomes invested with magnetic flux 

 in the manner represented in Fig. 7. Here the curves of magnetic 

 flux distribution, indicated by arrows, are circles eccentric to the 

 wires. One-half the energy of each moving block is electric and is 

 resident in the electric flux, and one-half is magnetic and is resident 

 in the magnetic flux. It will be observed that the magnetic flux is 

 so directed as to pass through the loop formed by the two wires, in 

 planes perpendicular to the wires. Moreover, the curves of mag- 

 netic flux stream-lines are all perpendicular to the curves of electric 

 flux stream-lines, which, already shown in Fig. 2, are here repre- 

 sented by dotted lines. The magnetic flux at each point is due to 

 the movement of the electric flux through the ether at that point, 



