goes in the opposite direction to the points, is merely the result of the continuous repulsion that occurs between the current that traverses the movable metal branch, and that which escapes out of it to penetrate into the air. Ampère's Theory on the Constitution of Magnets; and the After having studied the mutual action of electric currents upon each other, and having determined their laws, Ampère endeavoured to connect the action of currents and of magnets by means of a very ingenious hypothesis on the nature of magnetism. By carefully analysing the action of the different parts of a magnet upon a movable current, and that of a current upon the different parts of a movable magnet, he saw that these actions were exactly the same as those which might have occurred had the section of the acting magnet, or the magnet submitted to action, been replaced by an electric current circulating around this section and consequently closed, and situated in a plane perpendicular to the axis of the magnet. By observing that in some cases there was attraction and in others there was repulsion between the section of a magnet and an electric current, the direction of which he knew, he further succeeded in determining what the direction ought to be of these hypothetical currents; and for this he rested simply upon the law that there is attraction when the currents move in the same direction, and repulsion when they move in opposite directions. The following is the mode by which we succeed in attaining to this determination. We take a prismatic magnetised bar, using the precaution to hold it horizontally, so as to have the north pole on the left hand; we present it to the vertical branch of the movable astatic conductor, placed alone in the circuit; we find that, if this branch is traversed by the current proceeding upwards from below, it is repelled by all the parts of each of the faces of the bar, and this from one extremity to the other; that, on the contrary, it is attracted if its current is directed from above downwards. When thus passing each of the faces of the bar from one end to the other before the vertical current, we must take care always to hold the magnet horizontally, and the north pole to the left. Now if, by means of a little soft wax, we apply on each of the faces of the magnet small arrows of card, with the point turned in the direction according to which a current ought to travel in order to produce upon the movable conductor the attraction or repulsion that is determined in it by the action of a magnet (Fig. 93.), we find Fig. 93. that these small arrows represent a current circulating around each of the sections of the magnet, everywhere in the same direction, namely, from top to bottom in the face that is turned toward the movable conductors, and from bottom to top in that which is opposite to it, and moving from the conductor in its lower surface, and approaching it in its upper. The sum of these directions perfectly constitutes a current, circulating around each section of the magnet, as in a closed circuit. When, without changing the position of the arrows, we turn over the magnet, placing the north pole to the right, it is easy to comprehend that, as their direction is then inverted by the mere fact of this turning, it follows that the current which they represent goes from the bottom to the top in the face that is presented to the movable current. Thus, between the different parts of the magnet that are successively presented to the current, and the current itself, there is repulsion when the current is directed from above downwards, and attraction when it is directed from below upwards; actions precisely contrary to those that occurred in the preceding case, namely, before the magnet had been turned over. We obtain a perfectly similar result by presenting a horizontal current to the different sections of a magnet, Fig. 94. suspended vertically to a wire by one of its extremities (Fig. 94.), where we indicate, by small arrows fixed upon the different faces of the magnet, and at different heights, the direction that the currents ought to have which are supposed to circulate around its surface, in order to account for the attractive and repulsive effects that are observed. A magnet may therefore be considered as formed by an association of electric currents, all circulating in the same direction around its surface, and all situated in planes parallel to each other and perpendicular to the axis of the magnet. With regard to the direction of these currents, we have seen, by analysing the mutual action of a magnet and a current whose direction is known, that it is such that, if we hold the magnet horizontally before us, the north pole being to our left hand, the current goes from top to bottom in the exterior face most distant from the observer, and from bottom to top consequently in the face that is nearest or withinside. In order to fix this direction well in the memory, it is more convenient to suppose the magnet in its natural position, that is to say, in that position which is imparted to it, when it is movable, by the directive force of the earth, its north pole consequently turned towards the north; we then find that the direction of the arrows, which we will still leave in their place, is such as to indicate that the current is directed from the east to the west in the lower face of the magnet, and consequently from west to east in its upper face; that it is ascending in the face situated on the west, and descending in that which is on the east. We may further add, that it is evident that the form of the circuit in which each of these parallel currents circulates, whose association forms the magnet, depends on the exterior form of the magnet itself: that it is circular when the magnet is cylindrical; rectangular when its figure is that of a parallelopipedon; and that it is a series of rectangles, diminishing in size from the middle towards each of the extremities, when the form of the magnet is a lozenge. It is easy to see that Ampère's hypothesis of the constitution of magnets, as we have just expressed it, explains in the most satisfactory manner Oersted's fundamental experiment; as also all those that relate to the deviation of a magnet or a current, produced by the mutual action they exercise upon each other. All these effects may be traced to those that result from the mutual action of two currents upon each other—an action, by virtue of which they tend to place themselves parallel to each other, so as to be moving in the same direction. In order that this parallelism may occur with currents that circulate around the magnet, it is evidently necessary that the latter should find itself placed transversely to the current that acts upon it, or upon which it acts. Now, it is actually to this that the directive action tends, which is manifested in the experiments in which the movable current or magnet, instead of being able to obey attraction or repulsion, can only turn around a central point. In order to confirm the hypothesis to which he had been led of the nature of magnetism, Ampère endeavoured to arrange electric currents in the same manner as he had conceived they were naturally arranged in a magnet; and he thus succeeded in obtaining assemblages of currents which possess all the properties of true magnets. With this view he took a copper wire, and twisting it into a helix, taking care that the successive spirals did not touch each other, he brought back the two ends interiorly along the axis of the helix to its middle, and then, making the two extremities come out, without being in contact either with each other or with any part of the helix, he bent them so as to be able to suspend the whole as a movable conductor to an apparatus similar to that of Fig. 82. Then, having made the current pass through the wire of the movable conductor, he found that he had a true magnet, the axis of which and the two poles were the axis and the extremities of the helix (Fig. 95.). An ordinary magnetised bar ex ercised upon its extremities the same attractive and repulsive actions that it would have exercised upon those of a compass needle. In order to obtain more marked effects, it would be better to use, in the construction of the helix, a wire covered with silk; we may then bring the spirals of the helix close even to contact, without any fear of the current's passing directly from one to the other, instead of pursuing its course; for there is no metallic communication. In order to give greater firmness to the wire, the helix is wound around a |