Dynamo-Electric Machines. 293 



A could be got rid of by making n = v, which satisfies the con- 



X 



ditions. Then there remains a hyperbolic governor, pv x ~ 1 = B. 

 This does not, however, appear to be a good solution in prac- 

 tice. One would be disposed to have v at least three times as 

 great as n. In my Gramme machine as I have rewound it, 

 R/R + p is about 1/55, n/v about 1/2 : this is well within the 

 conditions, and yet past the point v = n. For practical purposes 

 the shunt excitation would bear to be much stronger. It is 

 difficult, however, to deal with practical applications at pre- 

 sent, as we know nothing definite about the governing func- 

 tions of the engines. 



As x approaches unity the sharpness of the governor re- 

 quired is much increased. This seems to point to a control of 

 the governor by the current ; for this increases as x approaches 

 1, or as the magnets approach the point of saturation. 



An interesting result is obtained in the case of main coils 

 only, where E is to be constant. 



Putting p = oo,v=0, 5 = 1, in the general formulae, 



KKTrfW, (1) 



E 2 

 R+r ; 



**£& w 



from (1), 



whence from (2), 



(l\v) x - 1 

 R + rsn^ — J T — 



E^ 



2a- 1 



E*- 1 



P= : 



n(l\v) x - 1 

 For x = 2, this is of the form 



A 



a hyperbolic governor of moderate sharpness. 



If #=3, P — ~V 



v* 



less sharp. If x diminish nearly to 1, the sharpness required 

 is greatly increased. 



These cases are interesting, since they explain the possi- 

 Phil. Mag. S. 5. Vol. 15. No. 94. April 1883. Y 



