of the current through both spirals were the same and when they were opposite, compels the conclusion, that, in the position above described, the wires exercise no influence upon each other, which result was indeed anticipated. The spirals were now set normally opposite, and brought within a line of each other. In the following experiments the combinations of the spirals were essentially the same as in the former; the observations differing merely therein, that for the first vertical column, under the head 'Temperature,' the thermometer stood nearest to the inner coating of the battery, while for the second column the spirals were nearest to the same coating.

Principal current in the same direction through both spirals.

A comparison of these results with those first obtained shows that an alteration of the current takes place when one portion of the connecting wire is brought near to another and parallel to it. In order to express the result in a brief manner, let the connecting wire be conceived to be of three different shapes; stretched out straight, bent into the form of a U, and bent into the form of an N (the two parallel sides of the latter being brought very near each other). The discharge of a battery through the N form gives the feeblest result, that through the U form the strongest, the current through the straight wire being intermediate between both. In general the modification of the current is so inconsiderable, that in common experiments with the battery it may

be entirely neglected. To place the fact beyond doubt, we see that it is necessary to place 107 feet of a current 119 feet in length within a line of each other, by winding them into two spirals, the most distant parts of each of these being not more than a foot apart. In the wire screws, used so frequently in the circuit for the sake of sparing room, the distance between two windings is much less than the diameter of the screw; hence a wire wound into this form must, as in the case of the N wire, principally weaken the current.

In the series of experiments given above, the spirals were united by a copper wire 29 inches long and ths of a line in thickness. The measured currents bore nearly the following proportions to each other:—

Without the action Current in same direction.
of the spirals. N combination of the spirals.
100
89

Current in opposite directions. U combination of the spirals. 106

The connecting wire was exchanged for one of 39 inches in length and th of a line thick. Out of twenty-seven observations, the following numbers were found for the respective currents:

100

91

106

Finally, a steel wire, 344 inches in length and 74ths of a line in thickness, was introduced, and with this I obtained the following numbers:

100

:

The phænomenon is therefore independent of the retarding value of the wire which unites the spirals; for here we have very different values of retardation, but currents of almost the same proportions. In all these cases the increase of the current is less than the diminution, which therefore must not be regarded as an accidental circumstance.

The result established may be thus expressed :

Two portions of the connecting wire of a battery, which run closely parallel, act upon each other. The current will be weakened by this action when its two parallel portions move in the same direction, and strengthened when they move in opposite directions.

The secondary current.-To excite a secondary current, two wires must be placed near each other; the most convenient way of effecting this being to wind them into spirals either as discs or cylinders (we shall call them in future induction-discs, or induction-cylinders). From former experiments the author was led to conclude, that the heating in the thermometer was proportional to the number of coils; but this is only approximately Phil. Mag. S. 4. Vol. 3. No. 17. March 1852.

N

correct, as the following experiments prove. It will add much to the reader's comfort if a clear conception of the arrangement of the spirals and thermometer be obtained. In the case now to be described there were two circuits, a primary and a secondary; in the primary circuit were the battery and two inductiondiscs placed one after the other, and in the secondary circuit two others of the same size, and the air-thermometer; the two primary spirals were connected by a wire which proceeded from the centre of one to the rim of the other, and the two secondaries were united in the same manner; the other ends of the secondary spirals were connected with the thermometer. The experiments were made in the following manner :- First, one secondary was placed parallel to its primary and two lines distant from it; the battery was discharged, and the power of the secondary current was observed on the thermometer; secondly, the other secondary spiral was brought within two lines of its primary, the two former being widely separated, and the heating was again observed; thirdly, both secondaries were brought within two lines of both primaries, and the strength of the current induced in the whole secondary circuit was ascertained: the following are the results of these experiments :

These experiments prove that the strength of the secondary current increases in a somewhat smaller ratio than the length of the wire, a result which might be predicted when the reaction of the secondary upon the primary current is taken into account. Thus, if we imagine the secondary circuit divided into two portions, the sum of the actions of both portions, taken separately, is greater than the action of both together.

The reaction of the secondary current upon the primary was demonstrated by former experiments in the following manner :The primary circuit contained two induction-discs, as in the case just described. Opposite to each of these was placed a secondary disc, which could be closed so as to form a continuous circuit in itself; the secondary discs were first left open, and the strength of the primary current was measured; the secondary discs were next closed by short copper wires, and the primary

current was again measured; the strength of the latter was found to be quite unchanged by this closing of the spirals. If, however, instead of being closed by a short copper wire, the ends of one of the secondary spirals were united by increasing lengths of fine platinum wire, the principal current was observed to decrease to a certain limit, from which forward it increased. A secondary current induced by a primary thus circumstanced may be expected to partake of the fluctuations of the latter, which conclusion has been established experimentally by the author in his present investigation. The primary circuit contained two induction-discs, and the secondary two others; the ends of one of the secondaries were united to the thermometer by short copper wire; the two ends of the other secondary were united, first by a short copper wire, and afterwards by increasing lengths of platinum wire of 0.028 of a line radius. The following table gives the result of the experiments:

Closing of the second spiral.

Copper 16inches. Platinum 1.98 feet. 5.95 17.9 37.6 97.2 Secondary current in the first spiral.

53 35 37 51

Here we observe that the effect of the platinum wire in weakening the principal current was a maximum when its length was 17.9 feet; and that the effect of a wire 5.95 feet in length was very nearly equal to that of a wire 97.2 feet in length.

The mediate action of one secondary current upon another is established by the foregoing experiments; for the secondary current in a spiral is shown to be modified by the action of another distant secondary spiral upon the principal current. But the direct action of one secondary upon another can also be shown by permitting both to be induced by the same portion of the primary circuit. The experiment is frequently so arranged, that, besides the spiral whose secondary current is observed, another spiral is introduced, either between the former and the primary circuit, or at the opposite side of the latter. Instead of spirals, metallic bodies have been sometimes introduced; and it has been found that the better conductor the body is, the greater will be the weakening of the secondary current; to this action Henry has applied the term "screening." The precise action which takes place will perhaps be best understood from the discussion of the next table, which contains the results of a number of experiments conducted in the following manner :-Round a wooden cylinder, 13 inches high and 6 inches diameter, three copper wires were wound spirally. Each wire was 7ths of a line in thickness, 53 feet long, and made 31 revolutions, possessing a 'pitch' of 4 lines. The first spiral was introduced into

the primary circuit, the second was connected with the thermometer, and the third was closed by wires of successively increasing retarding values. The results obtained are as follows:

When the third spiral was open, the value of the secondary current induced in the second was 100; the uniting of the ends of the third spiral by a copper wire 23 inches in length brought the secondary current down to 61. Now it is proved by experiment, that the action of a secondary spiral upon the primary current when its ends are connected by a wire of the above length (23 inches), is just the same as if the spiral were left open; hence the diminution of the secondary current is due, not to any modification which the primary has undergone, but to the direct action of the other secondary. The first four of these experiments show, moreover, that the stronger the current in the third spiral, the greater is the amount of weakening in the second*. By increasing the resistance in the third spiral, the primary current at length becomes modified, and a decrease of the current in the second spiral up to a certain point is the consequence; from this point forward the current again increases, until with a retarding value of 126459 it attains almost the same strength which it possessed when the third spiral was altogether inactive. The result of this immediate and, mediate action of the one secondary upon the other, is the occurrence of two maxima and two minima in the series of observations.

*Hence the formation of a secondary current is checked by permitting the primary wire to excite a second secondary at the same time; the reader will do well to remember this, as the circumstance is turned to account further on.