Dec. 30, 1875] 



NATURE 



177 



has had dissimilar poles. By meditating upon certain facts 

 connected with this subject, I concluded that, by passing a current 

 from one end to the other of a magnetized rod or wire having 

 similar poles at its two ends, the magnet would probably rotate, 

 and experiment has demonstrated that conclusion. 



Upon a thin wooden tube 15 centims. long and 7 taillims. 

 bore, I wound a cotton-covered copper wire i" 7 millim. diameter, 

 from one end of the tube to the middle, then reversed the 

 direction of winding, and continued to the other end and back to 

 the middle, again reversed, and coiled to the first end of the tube ; 

 by which arrangement the passage of a current through the coils 

 produced two similar poles at the ends of the tube, and two 

 others of the opposite kind at the middle. 



The tube being now fixed in a vertical position, a straight 'ron 

 wire 15 centims. long, and i*8 millim. diameter, pointed at its 

 lower end, and surmounted by a brass mercury cup 5 miilims. 

 diameter, containing a drop of mtrcury, was supported entirely 

 within the tube and free to rotate, by a similar cup (surmounting 

 a fixed vertical brass rod), at the lower end of the tube ; the 

 upper end of the axial wire being kept in position by a vertical 

 brass rod fixed above the coil and terminated at its lower end 

 by a sharp point of platinum in the mercury cup. 



A current from 6 one-pint Grove's elements, arranged as 3, 

 being now passed through the coil, brass rods, and axial wire, the 

 latter rotated rapidly. 



A copper wire substituted for the iron one would not rotate, 

 probably because copper is so little capable of acquiring longi- 

 tudinal magnetism. 



To ascertain if the coil-cnrrent simply performed the function 

 of longitudinally magnetizing the axial wire, I took an iron 

 wire 23 centimes, long and 27 miilims. diameter, sharp-pointed 

 at its lower end ; soldered to its upper end a double wire of 

 cotton-covered copper, each wire being 1 7 millim. diameter, 

 coiled the double wire upon the axial rod in two layers, and so 

 as to enable two similar poles to be formed at the extremities of 

 the axis, and terminated the copper wires by a little brass 

 mercury-cup just above the top end of the vertical iron axis. By 

 supporting this apparatus as the axial wire in the previous 

 experiment, and passing the current, rotation occurred. 



Keveriing the direction of the current did not reverse the 

 direction of rotation. 



These experiments, produce a striking effect in a lecture, be- 

 cause the rotation appears to be produced without reaction of the 

 moving pare of the apparatus upon any external or fixed body. 



In each of these cases of rotation, an upward vertical current 

 entering a lower south pole or leaving an upper one, caused the 

 upper end of the rod to rotate in the direction of the hands of a 

 watch, and a downward current entering or leaving a north 

 pole also produced that direction of motion, and reversing the 

 poles in either case reversed the effect. 



In each of these instacces of rotation, without the aid of a 

 current rear the middle of the magnet, the coil being so con- 

 structed that the current in it could not be reversed without 

 reversing that in the fixed conductors near it, reversing the 

 direction of the current did not reverse that of the rotation, 

 because the two acting influences were reversed together, and 

 therefore each apparatus had its own direction o! rotation, 

 either right handed ( V^) or left handed, * according to the direc- 

 tion in which its coils were wound. It follows from this that 

 a current, the direction of which is alternately reversed, will 

 drive the apparatus quite as well as one in one uniform direction. 



I now endeavour to iricrease the effect. Foi this purpose I 

 ■.bstituted for the upper brass rod a fixed coil consisting of one 

 .yer of copper wire upon an iron wire axis, but having dis- 

 ;milar poles at its ends and no poles at its middle part, and 



laced between it and the lower brass rod a right-handed one 

 ;;ee to rotate. The opposed poles of the fixeS and m.ovable 

 coils were of opposite kinds, /.IT. north and south. On passing 

 a current from a Noe's thermo-pile of 96 elements, + connected 

 as 24, rapid rotation in a right-handed direction occurred. I 

 now substituted for the lower brass rod another fixed coil, 

 similar to the upper one, but of an opposite direction of polarity, 

 and passed the current again ; still more rapid rotation in the 

 same direction took place, and the effect was very striking. 

 In this latier instance, two south poles free to move were opposed 

 to two fixed north poles, and in each instance the current was 

 passed upwards. 



I now substituted for the movable coil a vertical wire of iron 



• By a "right-handed " coil, I mean one the upper end of which rotates 

 in the same direction as the hands of a watch. 



t I have found this apparatus very convenient for such experiments. 



13 centims. long and 17 millim. diameter, surmounted by a 

 small brass mercury cup ; passed the current from the thermo- 

 pile, and obtained rotation, but less rapid than before ; but by 

 inclosing this wire in the axis of a fixed coil which produced, 

 appropriate and similar poles at its two ends, as in paragraph 3, 

 and repeating ihe experiment, very great velocity of rotation was 

 obtained. Rotation of a somewhat thicker wire of nickel was 

 also obtained, both with and without the aid of the current in 

 the middle fixed coil. I also tried, without the aid of the middle 

 fixed coil, and with it, a copper wire of similar dimensions to 

 the iron one, and obtained rotation freely. Each of these 

 rotations agreed in direction with those of the movable coil. 



The apparatus represented in the annexed sketch was em- 

 ployed for nearly all the various modifications of the experiment, 

 by substituting for one or more of the coils metallic wires, &c., as 

 desired. The upper part of the brass pillar A was capable of 

 sliding in the lower part B, and could be fixed by a screw C, 

 which encircled the spht end of the tube B. The fine adjust- 

 ment was effected bv means of the screw D, the lower end of 

 which rested upon the top of a tall brass rod inside the brass 

 pillar. The upper and lower fixed coils or rods E and F were 

 insulated from the brass clips G and H, and the battery was 

 attached to the binding-screws I and J. K is a binding-screw 

 for connecting with the upper coil or rod. 



I also obtained rotation of the iron wire whilst the wire was in 

 a horizontal position, its ends resting in hollows in the ends of 

 the iron axes of the two fixed coils, and the ends of those axes 

 and of the movable wire lying upon the surface of pools of 

 mercury in small watch-glasses. The movable iron wire was 

 inclosed in the axis of a thin iron tube w ithin a fixed coil, having 

 appropriate and similar poles at its ends. The current from the 

 thermo-pile produced very rapid rotation. This result proves 

 that the rotations are not due to terrestrial magnetic influence. 



As the directions of magnetic polarity, electric current, and 

 rotation agree with those in the different forms of Ampere's 

 experiment, and as in most, if not all, of the previously known 

 cases of rotation of a bar-magnet or conducting-wire on its axis 

 an electric current passes through the end of the bar or wire, it 

 is evident that those rotations were due, not only to the portions 

 of current in the mercury, and fixed conductors connected with 

 it, near the middle of the magnet or wire, but also to the 

 influence of the currents in the fixed conductors near the ends 

 of the magnet or wire. 



\_Noie added September, 1875. — It having been suggested by 

 Professors Maxwell and Stokes that the rotation in the foregoing 

 experiments was due to the influence either of the magnetism 

 of the fixed magnets or of the current in the fixed conductors, 

 near the ends of the movable wire or magnet, upon the portions 

 of current in the cups of mercury, I diminished the internal 

 diameter, both of the upper and lower cups, from 4 miilims. to 

 175 millim., and arranged the following apparatus and experi- 

 ment. 



The fixed upper wire was of brass 2*5 miilims. diameter and 

 60 miilims. long ; it had no coil upon it, and was used as a 

 conductor only ; i^s lower end terminated in a fine point of a 

 steel needle protecting 6 miUims. The lower fixed wire, also 

 used as a conductor only, was of platinum to resist the action of 

 the mercury ; it was 2 3 miilims. diameter and 75 miilims. long, 

 with a cavity in i's upper end 35 miilims. deep and 1.75 milUm. 

 diameter, and containing a thin plate of ruby in its lower part, 

 with a minute hole in the centre for the ne^cle point to rotate 

 in. The movable wire was 2 '5 miUims. diameter and 125 

 miilims. long, is upper half being composed of soft iron and its 

 lower half of brass ; its lower end terminated in a needle-point 

 like that of the upper fixed wire, and its upper end had a cavity 

 and perforated ruby plate hke that in the lower fixed wire. A 

 voltaic coil 60 miilims. long and 7 miilims. internal diameter, 

 composed of four layers of cotton covered with stout copper 

 wire, was used to magnetize the iron half of the movable \^-ire, 

 and fixed by means of a separate support in a proper vertical 

 position beforehand, so as to inclose in its axis the iron wire 

 portion only. The little cups were also each half fiUed with a 

 minute globule of mercury before putting the movable wire 

 into its place. 



After adjusting the wire so as to make rotation easy, a currert 

 from 6 Grove's elements of cne-pint capacity, arranged as a 

 series of 6, also as a double series of 3, was passed through 

 the coil and vertical wires ; and the direction of the portion of 

 the current in the coil alone, also in the vertical wires alone, 

 was varied ; but notwithstanding that plenty of current passed, 

 no signs of rotation cotild be detected. These results, therefore. 



