POWER DEVELOPMENT AT NIAGARA. 627 



high voltage on account of the very much more perfect insulation 

 possible. 



The advantage, of course, of using a very high electrical pres- 

 sure lies in the principle stated above of the loss in sending a 

 given amount of energy over a given wire being inversely pro- 

 portional to the square of the voltage. 



By the use of step-up transformers it will, of course, be pos- 

 sible to transmit at any voltage that the insulation of the line 

 can withstand ; but if this high voltage could be reached by the 

 machines directly, the loss (we may liken it to a friction loss in 

 machinery) of efficiency in the transformers, and, even more im- 

 portant, the great cost of that part of the equipment, would both 

 be avoided. 



What will be done will be to use these step-up transformers and 

 put current on the transmitting line at about twenty thousand 

 volts ; it is likely, however, that in any subsequent enlargements 

 of the generating plant the three original machines will be used 

 for local work only, and a radical change made in the direction of 

 an enormously higher generated voltage. 



Intimately associated with this question is the problem of how 

 to convey current at this tremendous potential of twenty thou- 

 sand volts to distances. An idea of what it means may be had 

 from the facts that two thousand is relied on to be sufficient to 

 instantly kill a human being, and that the energy of a current 

 given up in passing through any given resistance varies as the 

 square of the voltage. 



The chief difficulty to be met in such line construction is that 

 of efficiently insulating the wires. If one attempted to use a line 

 insulated merely as an ordinary telegraph line is, there would 

 be an enormous loss, amounting practically to the whole of the 

 transmitted current, in moist weather, by leakage over the damp 

 surfaces of the glass or other insulators. The remedy for this 

 leakage would, however, be a comparatively simple matter by 

 means of well-known oil-holding arrangements for the insulators 

 were it not for the further fact that it is imperatively necessary 

 not to have the two wires, the going and return ones, farther apart 

 than can not be avoided on account of what are known as the 

 effects of self-induction. The wires strung on telegraph poles 

 would have to be so far apart in order to insure their never, by 

 any possibility, coming in contact, that the self-induction losses 

 would make that method impracticable. 



The self-induction of a circuit has the effect of retarding both 

 the starting up and the dying out of a current flowing in the cir- 

 cuit. The phenomenon gives a resemblance of the possession of a 

 property analogous to mechanical inertia to the current. Since 

 inertia, however, is a property dependent solely on the mass of a 



