It is doubtless well known that if, as in Fig. 1, a Gramme ring armature is connected to leads at four points as shown and a magnet is revolved inside of it (or if the ring is revolved in a magnetic field and the current led off by contact rings instead of a commutator), there will be two alternating currents generated, which will differ from each other in their phases only. When one is at a maximum the other is zero. When such a double current is sent into a similarly constructed motor it will produce or generate what might be called a rotary field, which is shown diagrammatically in the six successive positions in Fig. 2. The winding here is slightly different, but it amounts to the same thing as far as we are concerned at present. This is what Mr. Dobrowolsky calls an "elementary" or "simply" rotary current, as used in the Tesla motors. A similar system, but having three different currents instead of two, is the one used in the Lauffen transmission experiment referred to above. [Illustration: FIG. 1.] [Illustration: FIG. 2.] In investigating this subject Mr. Dobrowolsky found that the best theoretical indications for such a system would be a large number of circuits instead of only two or three, each differing from the next one by only a small portion of a wave length; the larger their number the better theoretically. The reason is that with a few currents the resulting magnetism generated in the motor by these currents will pulsate considerably, as shown in Fig. 3, in which the two full lines show the currents differing by 90 degrees. The dotted line above these shows how much the resulting magnetism will pulsate. With two such currents this variation in magnetism will be about 40 degrees above its lowest value. Now, such a variation in the field is undesirable, as it produces objectionable induction effects, and it has the evil effect of interfering with the starting of the motor loaded, besides affecting the torque considerably if the speed should fall slightly below that for synchronism. A perfect motor should not have these faults, and it is designed to obviate them by striving to obtain a revolving field in which the magnetism is as nearly constant as possible. [Illustration: FIG. 3.]