Page:Scientific Memoirs, Vol. 1 (1837).djvu/628

616 {| align=center Hence may be calculated by means of the least squares therefore we have for the calculated values of $$\alpha$$
 * rowspan=2|Number of Convolutions. || colspan=4|Deviations. || rowspan=2 |Mean Deviations or $$\alpha$$.
 * | 1 ||| 2 ||| 3 ||| 4
 * | 5 ||| 8·6 ||| 8·7 ||| 8·5 ||| 8·6 || 8·63
 * |10 || |17·5 || |17·8 ||| 17·2 || | 17·1 || 17·40
 * | 15 ||| 26·4 ||| 27·2 ||| 26·6 || |25·6 || 26·45
 * |20 ||| 35·5 ||| 35·3 ||| 35·6 || |34·6 || 35·25
 * |25 ||| 45·2 ||| 46·0 ||| 45·0 || |44·2 || 45·10
 * 30 || |54·6 || |56·5 ||| 55·0 || |54·1 || 55·05
 * }
 * | 15 ||| 26·4 ||| 27·2 ||| 26·6 || |25·6 || 26·45
 * |20 ||| 35·5 ||| 35·3 ||| 35·6 || |34·6 || 35·25
 * |25 ||| 45·2 ||| 46·0 ||| 45·0 || |44·2 || 45·10
 * 30 || |54·6 || |56·5 ||| 55·0 || |54·1 || 55·05
 * }
 * 30 || |54·6 || |56·5 ||| 55·0 || |54·1 || 55·05
 * }
 * }

Here then the coincidence for this kind of experiments is very great, go that we may regard the position as entirely confirmed, namely that

Moreover, we must not let it escape our attention, that in all the three series of observations the differences of the calculated and of the observed deviations are in the beginning positive, and then negative; which seems to show that the electromotive power increases in a somewhat quicker proportion than the number of the convolutions; but the differences are so small, and become, when the observations are made with great care (as the third series proves) smaller and smaller, I therefore ascribe this little irregularity to the influence of some peculiar circumstance which up to the present moment I have not succeeded in discovering.

In these experiments I employed at first the horseshoe magnet, but