534 
MR. T. GRAY OR THE MEASHREMERT OF 
gate the effect of cyclic speed on the dissipation of energy, using a varying impressed 
electromotive force. 
The results here quoted were obtained from experiments on a large electromagnet 
belonging to the Electrical Laboratory of the Rose Polytechnic Institute. I have, 
however, found it easy to trace curves for an ordinary 6000-watt transformer, 
although the time is in this case considerably shorter than thad required for the 
magnet. I hope to supplement the results here given by results for various trans¬ 
formers, as soon as they can be reduced. The electromagnet was constructed in the 
shops of the Ptose Polytechnic Institute, and was designed by my colleague. Dr. C. 
L. Mees, and myself for convenient use in the instruction of students in magnetic 
measurements. It consists of two cores C C (fig. 2), each surrounded by four coils of 
Fig. 2. 
insulated wire, 2'5 millims. diameter. Each coil is divided into four concentric coils 
of five layers each, and there are 24 turns in each layer. The total number of turns 
in the magnetizing coils, when all are joined in series, is thus 3840, while the arrange¬ 
ment allows a considerable variation to be made, if desired, in the distribution, position, 
power, or resistance of the coil. For example, the inside set of concentric coils may 
be compared with the outside set to illustrate the law that the magnetizing effect is 
proportional to the number of turns, and for closed magnetic circuits practically 
independent of the diameter of the turns. Again, one set of alternate coils may be 
used for a primary, wlnle the others, joined in any way, may be used for the secondary 
of a transformer. The magnet, when joined in this way, is capable of transforming 
from 30,000 to 40,000 watts. For transformer purposes this arrangement has the 
advantage for purposes of instruction and experiment that the action of an open and 
a closed circuit transformer can be directly compared. Various other modifications 
of the arrangement of the magnet coils and cores will readily be seen, with which we 
are not concerned in this j^aper. 
The cores C C rest on a yoke Y, and on the top of the cores the pole pieces P P 
rest. The lengtli of the iron circuit can be varied by moving the cores further 
apart and pushing forward the pole pieces. To facilitate this operation, a screwed 
rod is run from end to end of the yoke, sunk just beneath the surface; one half has 
a right, and the other half a left-handed screw, which work in half nuts fixed in the 
