would necessarily increase the force of the electrical current, I therefore made the following experiment. I wound round the armature ten convolutions of copper wire bespun with silk, and the conducting wires were connected with the ends of this spiral only by a single twist of the wires; the result of the four readings off amounted to 36°·8; upon this the same connection was made by twisting the ends of the wires ten times round one another as tightly as possible; the deviation amounted again to 36°·8; I finally pressed the last connection as tightly as possible together with a pair of pinchers, so that they were very much flattened; the deviation was 36°75. We may therefore consider the connection made by tightly twisting the wires ten times round one another as quite sufficient, and this was therefore made use of in all the subsequent experiments. The places where the connection was made were then wound round with silk stuff in order to secure them from reciprocal contact.
The second preparatory experiment I made in order to see whether, when I advanced the electromotive spiral on the armature more to the north limb or to the south limb of the magnet, it had any influence on the electric current. For this purpose I obtained with two convolutions the following results:
| The convolutions advanced till in contact with the north limb of the magnet, gave a deviation | = | 5°·55 |
| The convolutions advanced until in contact with the south limb of the magnet, gave a deviation | = | 5°·55 |
| The convolutions advanced to the middle of both limbs gave a deviation | = | 5°·60 |
therefore this influence also of the different positions of the spirals on the armature is imperceptible: from this time I always placed them so that the spirals occupied the middle of the armature.
I thirdly determined, before I proceeded to the proposed experiments, the thickness of the copper wires employed; I weighed two feet of each having wound off the silk, by which I obtained the proportions of their diameter on which it principally depended; but in order to obtain also their absolute thickness, I measured the thickest by means of a micrometrical contrivance: I obtained the following results, in which I have designated the wires, beginning with the thinnest, Nos. 1, 2, 3, and 4.
| grains | inch. | |||||||
| 2 | feet of wire | No. 1 | weighed | = | 23·3; | absolute thickness | = | 0·023 |
| 2 | No. 2 | = | 27·4; | = | 0·025 | |||
| (wire of the multiplier) | ||||||||
| 2 | No. 3 | weighed | = | 83·9; | absolute thickness | = | 0·044 | |
| 2 | No. 4 | = | 166·1; | = | 0·061 | |||
All the four kinds of wires were well covered with silk, so that no
