the Goldsclimidt Dynamo. 759 



(2) The second point is that an armature revolving in 

 front o£ such a field magnet, which i£ the magnet were 

 steady would have the frequency q generated in it, will, when 

 the field magnet itself is excited by an alternating current of 

 frequency p, be subject to the resultant frequencies p — q and 



p + q» 



For if the armature stood still, the induced frequency 

 would be p ; and if it revolves, its own frequency q is super- 

 loosed upon this ; in one direction running after and perhaps 

 overtaking the frequency p, and in the other direction 

 running away from and thereby virtually increasing the 

 relative magnetic velocity. 



If the two frequencies are equal, being both generated by 

 the highest rate of rotation mechanically safe, the two resultant 

 frequencies will be and 2p. 



This is the principle which it was proposed to utilise in the 

 series of machines mentioned above as under discussion in 

 1892. 



(3) This is not real doubling, but an addition of one to 

 one, as can be proved hj continuing the process and employing 

 the frequency 2p as again a field magnet exciter; for, when 

 an armature is revolving in front of that, the frequencies 

 generated in that new rotor are Zp and p respectively (the 

 sum and difference). Hence the progression is merely 

 arithmetical, not geometrical. If, still further, the 3p is 

 used as exciter, the frequencies appearing in a new rotor will 

 be 4zp and 2p ; and so on. 



(4) The frequency in the first rotor is of course not a 

 current at all — though it sounds like a continuous current — 

 but it is static magnetism, and may be said to correspond 

 with the natural reaction between field magnet and armature, 

 like that between a magnet and its keeper, when both are 

 stationary. 



(5) The 2p generated in the second armature as the 

 difference between stator and rotor frequencies is likely to 

 be opposite in phase to the 2p generated in the same armature 

 as the sum of rotor and stator frequencies ; and the two may 

 therefore partially cancel each other. Similarly with other 

 intermediate frequencies in the chain, each of which occurs 

 twice over. The only unique currents will be the first and 

 the last — the steady current in the magnets of the original 

 exciter, and the highest frequency currents in stator and in 

 rotor at the other end of the chain. 



(6) The self-induction of the different coils, which would 

 naturally cause lag and entail considerable complication and 

 impedance, can be neutralised for any desired frequency by 



3F2 



