regret that I shall not be able to see it any more, for I should have liked to try to make it act in an opposite direction, that is to say, to produce a current or an electric light by means of mechanical work.”  A little more than two years later these experiments were carried out on a larger motor constructed by Kravogl in 1869, and Mr. Pfaundler was enabled to write as follows:  “Upon running the machine by hand we obtain a current whose energy is that of one Bunsen element.”  This letter is dated February 11, 1870, that is to say, it is a year anterior to the note of Gramme.

[Illustration:  FIG. 1.]

In the presence of the historic interest that attaches to the question, we do not think it will be out of place to reproduce here the considerations that guided Prof.  Pfaundler in the researches that led him to convert the Kravogl motor into a dynamo-electric machine.  Let us consider two magnetized bars, db and bd’, placed end to end and surrounded by a cylindrical armature forming a shell, this armature being likewise supposed to be a permanent magnet and to present poles of contrary direction opposite the poles of the bars.  For the sake of greater simplicity this shell is represented by a part only in the figure, s n n s.  If, into a magnetic field thus formed, we pass a spiral from left to right, the spiral will be traversed by a current whose direction will change according to the way in which the moving is done.  It is only necessary to apply Lenz’s law to see that a reversal of the currents will occur at the points, a and c, the direction of the current being represented by arrows in the figure.  If we suppose a continual displacement of the spirals from left to right, we shall collect a continuous current by placing two rubbers at a and c.  Either the core or the shell may be replaced by a piece of soft iron.  In such a case this piece will move with the spiral and keep its poles that are developed by induction fixed in space.  From this, in order to reach a dynamo-electric machine it is necessary to try to develop the energy of the magnetic field by the action of the current itself.  If we suppose the core to be of soft iron, and make a closer study of the action of the current as regards the polarity that occurs under the influence of the poles, s, n, s, we shall see that from d to a and from b to c the current is contrary, while that from a to b and from c to d’ it is favorable to the development of such polarity.  In short, with a spiral moving from d to d’ the resulting effect is nil, a fact, moreover, that is self-evident.  Under such circumstances, if we suppose the shell, as well as the core, to be of soft iron, we shall obtain a feeble current due to the presence of remanent magnetism; but this magnetism will not be able to continue increasing under the influence of the current.  To solve this difficulty