magnets in series. If it be absolutely necessary to use a motor of that class, the field magnets of which are either permanent steel magnets or are electro-magnets excited independently, the waste can to a certain extent be prevented by inserting into the armature current an electromagnet which will by its self-induction steady the current. Since this point is of importance, the author has thought it necessary to verify the above theory by experiments. These were undertaken with a twofold object. First, to prove that in a series-wound motor there is no appreciable waste of current at the dead points, and, secondly, to prove that in a motor the field magnets of which are separately excited, such waste occurs. The experiments were carried out as follows. Two small Griscom motors were placed in line behind each other, and their spindles were coupled, so that the armatures stood at right angles to each other, that is to say, when one armature was at its dead point the other was in the position of best action, and its counter-electro-motive force was a maximum. This disposition is represented in Fig. 15 by the dotted curve overlapping that shown in full lines by 90°. The resultant counter-electro-motive force is at any point the sum of the ordinates of the two curves, and is shown by the undulating line a b. It will be seen that this curve nowhere touches the horizontal and, therefore, the total counter-electro-motive force of the two motors coupled in series never is zero. An abnormal rush of current at the dead points of any of the armatures can, therefore, not take place. The motors were supplied with a current, the electro-motive force of which was kept as nearly as possible constant during each experiment, whilst the mechanical energy developed was measured on one of the author's absorption dynamometers. The commercial effi

EXPERIMENTS ON SELF-INDUCTION.

65

ciency of the two motors combined was thus ascertain ed, as shown in Table I. The motors were then coupled parallel, and their efficiency was determined under the same conditions. In this case there were, during each revolution, four dead points, at which the counterelectro-motive force was zero, and when an abnormal rush of current could take place if not checked by the self-induction of the magnet coils. As was to be expected, the current passing through both motors was about double, and its electro-motive force was about half of the former values. But the commercial efficiency was about the same, Table II. One motor alone was then tried, and its commercial efficiency was found to be about the same as that of the two motors combined, Table III. The field magnets of both motors were then excited separately, and the armatures coupled at right angles and connected in series, as per Fig. 15, when the commercial efficiency was found to be rather higher than in the former experiments, Table IV. This is but natural, because the energy necessary to excite the field magnets was not taken into account when calculating the efficiency. The two armatures were then coupled parallel-field magnets still independently excited-and thus during each revolution there were four points where the counterelectro-motive force was zero and waste of current did take place, as is clearly shown by the low efficiency in Table V. One motor alone was then tried under the same conditions and the same result was found, Table VI. These experiments prove conclusively that our above reasoning about the effects of the dead points is correct.

F

Test of Two Griscom Motors, Numbers 1017 and 1027.

Table I. Armatures Coupled at Right Angles, both Field Magnets and Armatures connected in Series.

Table II. Armatures Coupled at Right Angles. Each Armature in Series with its Field. Both Motors connected

EXPERIMENTS ON SELF-INDUCTION.

67

Table III. One Motor only. Armature and Field Magnets connected in Series.

Table IV. Armatures coupled at Right Angles and Connected in Series. Field Magnets excited separately.

Table V. Armatures Coupled at Right Angles and Connected in Parallel. Field Magnets excited separately.

Table VI. One Motor only. Field Magnets Excited

As already mentioned, motors with ordinary shuttlewound armatures have the disadvantage of requiring to be started by hand if they happen to have stopped on a dead point. They are, consequently, only made of small size, and for larger motor armatures without dead points are used. Such an armature can be evolved out of the simple shuttle-wound pattern by employing two sets of coils placed at right angles to each other. This arrangement is shown in Fig. 18, which represents the Hefner-Alteneck winding invented in 1872. In order to avoid complication the shaft is omitted and the core is indicated by two dotted circles. From what has already been explained it will be seen that in all those wires which at a given moment lie on the right hand side of the vertical centre line, the electro-motive force is directed towards the observer, and in all the wires lying to the left of that line it is directed from the observer. The diameter of commutation joining the points of contact of the brushes with the commutator cylinder will, therefore, be horizontal. In the position shown the negative, or left brush, will touch segment D, and the right or positive brush will touch segment B. The current enters the armature at the negative brush and splits into two circuits as follows:-One portion goes through VII.,