ALTERNATE CURRENT DYNAMO-ELECTRIC MACHINES 
231 
the electromotive force of the machine is modified, and that terms in 67 t£/T are 
introduced. The former may have real practical importance, and it is one of the 
objects of the present paper to ascertain how far it exists and is of importance in 
actual machines. 
Returning to machines as ordinarily constructed, in these the current in the 
magnet coils is not compelled to be constant, and any rapid variation of the induction 
in the magnet cores will induce currents in those cores. The variations in the current 
in the magnet coils and the currents in the cores both tend to annul the variations 
in the induction in the core arising from the current in the armature, but they will 
not tend to alter the average effect of the currents in the armature on the induction 
through the magnets. That there will be such an average effect is not difficult to 
see. Suppose the armature coils to be fixed in line with the magnets of the machine, 
any current in the armature will have its full effect in increasing or diminishing the 
field through the magnets. Suppose the armature coils to be fixed midway between 
the magnets, any current in the armature will then have practically little or no 
effect in increasing or diminishing the field through the magnets. If the armature 
be connected through resistance, inductive or non-inductive, and the machine run in 
the ordinary way, the current in the armature will lag in phase behind the electro¬ 
motive force E. The result is that when the armature is opposite to the magnets 
there is a current in the armature tending to demagnetize the magnets, and adding 
together the effects of the armature in all positions, there is an average effect tending 
to demagnetize the magnets. If the machine had a constant current round the 
magnets and divided iron in the magnets, this average effect, as well as its variations, 
would be fully accounted for by the term (Lir) 1 , call it self-induction or call it 
armature reaction, as you please. But inasmuch as the variations are in part annulled 
by the variations of current in the magnet-winding and the local currents in the 
magnet cores, we have a part of the diminution of the electromotive force E of the 
machine unaccompanied by retardation of phase of the current in the armature. 
Before giving any experimental results, it will be well to describe the machines 
used and the general method of experiment adopted. 
The two dynamos experimented upon were constructed by Messrs. Siemens Bros., 
and are of the same pattern and size but are of an old type. They are mounted 
upon a common base-plate, their pulleys being provided with flanges and bolts, so 
that any desired phase difference can be given to the armatures, for the accurate 
setting of which a graduated circle is provided, or so that the armatures can be run 
independently of each other. The pulleys have each a diameter of 12 inches, and 
are suited for a 6 -inch belt—the shaft is prolonged for the purpose of carrying a 
revolving contact-maker and a small pulley for driving a Buss tachometer. 
Each dynamo has a series of 24 electromagnets (see fig. 1 ), there being 1 2 on either 
side of the armature. The core of each magnet (A.) is of wrought-iron 2 y^ inches 
diameter, and 6 ^ inches long: and is wound with 5 layers, 35 convolutions per layer, 
