Study of Transformers. Ill 



using the above values for almost no loads and for great 

 loads. 



4th. If we imagine no magnetic leakage, 1=0, and we 

 have 



y v 



c= — ffT or= — p", (7) 



R + yTgR' R+-qR 



for all kinds of current variation unless R' is large. 



M 

 5th. We see that, except when R f is great, A'=pA, 



and when R' is infinity, A' = 0, and A = ^y ; so that, except 



for a very small range of small loads, we might expect the law 



. a I/A' 



to be nearly true. 



This law is very easily tested by means of the Tables. 



Having seen, then, from the Tables the small inaccuracies 

 of such a law, we might expect that generally 



Effective G=m + nX effective C l5 .... (8) 



where n is constant and nearly equal to p , however the fre- 

 quency &c. may alter, but m is a small constant which alters 

 if the frequency alters. We should expect this law to be 

 true when the secondary circuit is open, and also when the 

 secondary current changes from small loads to the very 

 greatest loads, and when it is short-circuited ; but for very 

 small loads it is somewhat untrue. For all practical purposes 

 it is true. 



6th. As, unless when R' is very great, 



S 

 A / = ^ M "P 



If R' = R + p, then 



S o 

 a = a P' ~-gi (9) 



This is Dr. Hopkinson's rule for the drop in the secondary 

 volts as the load increases, when the currents are sine func- 

 tions of the time. An examination of my Tables will show 



Phil. Mag. 8. 5. Vol. 32. No. 195. August 1891. N 



