'Electro- Magnetic Apparatus, 89 



most advantageous direction, yet as these forces are oblique, the re- 

 sultant is not equal to the sum of the components. Some of the fore- 

 going results, however, suggest a method of approximating them al- 

 most indefinitely. As an electrical current at right angles to the axis 

 of a magnet is known to increase its magnetism, and one parallel to 

 the axis to diminish it, the problem resolves itself into this; to cause 

 the electricity to enter, in equal quantities at all points of one extremi- 

 ty of the sheet, and to pass out in a similar manner at the other ex- 

 tremity, and also to exert the least magnetic effect possible, in pass- 

 ing to and from these points respectively, in directions parallel to the 

 axis. Now, from a comparison of some of the results which have 

 been stated in this paper, we may infer that there is a vast dispro- 

 portion between the magnetizing effect of an electric current, moving 

 near the temporary magnet, by means of a broad sheet of metal, and 

 the effect of the same quantity of electricity conveyed in an equal por- 

 tion of time, and at the same distance from the magnet, by a wire or 

 rod, of a length equal to that of the sheet, and of a size adequate to 

 the supply of the requisite quantity of electricity Xo the sheet, when 

 rods or wires are used merely as conductors between the extremities 

 of the sheet and the battery, and not for producing electro-magnetic 

 effects directly. This principle suggests the employment of a single 

 cylindrical rod at each end of the sheet, for establishing a communica- 

 tion between the latter and the wires of the battery, each rod being 

 parallel to the magnetic axis, and connected with the end of the sheet, 

 either continuously or by numerous points or wires, the connecting 

 parts being in either case so graduated in thickness, as to compensate 

 for the variable intensity of the electricity in different sections of the 

 rod, and thus transmit equal quantities to equal portions of one ex- 

 tremity of the sheet, and also equal quantities from equal portions of 

 the other extremity of the sheet to the rod. The magnet itself shoidd 

 constitute the rod at the interior extremity, in order to avoid the re- 

 moval of the current to an unnecessary distance.* For this purpose 

 a solid cylinder of iron, would be preferable to a hollow one. I have 

 shown the practicability of making the magnet itself a part of the cir- 

 cuit. The proper place for the direct attachment of the wires to a 

 horse shoe magnet would be on each side of the middle of the arch. 



* The only alteration necessary in the annexed drawing, if we wished to repre- 

 sent the apparatus with this improvement, would be a ridge on the roll to show the 

 brass rod to which //are attached. 



Vol. XXL— No. 1. 12 



