for the Distribution of Electric Energy. 193 
should be nearly constant. Various means may be resorted 
to for the purpose of securing this ; but we do not propose to 
go further into the matter at present than to investigate what 
the maximum difference of E.M.F. will be between different 
parts of the systems when the conductors are of the size which 
we should be inclined, from economical considerations, to adopt. 
Let the E.M.F. between the conductors at any point P of 
the system (say at the poles of the generator) be kept at the 
constant value e volts. Then, if the current in amperes per 
square centimetre section of the conductor be c, we have 
C = 10 9 |, 
where x is the E.M.F. in volts per unit length along the con- 
ductor. Now x is half the variation of E.M.F. per unit length 
along the system; and hence, if v be this variation, 
v=^>C$ (15) 
Let us assume, as before, that the maximum value of c is 
100, and we get 
2x100x1700 34 /1P , 
■= W s =10B ' ' • • ( 16 ) 
If we assume that the greatest percentage variation of E.M.F. 
allowable is p, we get for the distance d of the furthest point of 
the system from the point P, 
1000 
34 
d— —n—ep 
= 30ep (17) 
When p = 5, which produces a very inconvenient variation 
in the illuminating-power of the incandescent lamps at present 
made, we get with 
e= 50, d= 75 metres, 
e=100, d=150 » 
e=200, cZ=300 » 
and 
so on. 
Phil. Mag. S. 5. Vol. 16. No. 99. Sept. 1883. Q 
