were always interrupted by some accident, and I think that the action would otherwise have lasted still longer. The disengagement of gas was very inconsiderable, and took place only at the surface of the negative plate. The velocity was always at first from 120 to 122 revolutions in a minute, and decreased about half an hour after to 62 revolutions; a circumstance which is attributable to the commutator, which had not then the present construction. During the rest of the time the motion of the apparatus was remarkably uniform, making from 58 to 62 revolutions in a minute. I must however confess that I have obtained so extraordinary an effect only three times. There were always external circumstances, dependent upon the form of the voltaic apparatus, which counteracted the effect. I might be able to master most of these circumstances by constructing a new apparatus, the manipulation of which will be more convenient and the effect more certain.
14.
We have expressed by the magnetic force of each section of a wire traversed by an electric current. This force is measured by the deviation of the needle or by the magnetizing power of the connecting wire. By adopting the law of Faraday we may equally measure this current by the disengagement of the gas, which represents at the same time the cost of maintaining in action a voltaic apparatus. If be the quantity disengaged, we shall have . From this it follows that recurring to the formulæ of article 8, the œconomical effect may be expressed by the magnetic power of the whole extent of the connecting wire, divided by the development of the gas. This effect is in no respect changed either by the enlargement of the surface, or by the employment of various branches wound spirally around different bars of the same dimension. But by multiplying the helices, and uniting them to form a continuous wire, the œconomical effect may be increased as much as we please. For the disengagement of the gas, in employing helices or units of length, will be expressed by ; but we may put in action the magnetizing power of the whole extent of the conducting wire, and we shall have for the total force or When the magnetic bars are intended to produce a mechanical motion, the increase of the œconomical effect will reach its limit; since by multiplying the number of the bars, the weight of the apparatus and the friction of the pivots in the sockets will be at the same time increased, so that that effect can only be expressed by . The maximum of the œconomical
