G H

g h, which is separately insulated. the pins e, f rise out of the back of the fixed plates a and c, at unequal distances from the axis. the piece k is parallel to g h, and both of them are furnished at their ends with small pieces of flexible wire that they may touch the pins e, f in certain points of their revolution. from the brass piece m there stands out a pin i, to touch against a small flexible wire or spring which projects sideways from the rotating plate b when it comes opposite a. the wires are so adjusted by bending that b, at the moment when it is opposite a, communicates with the ball d, and a communicates with c through gh; and half a revolution later c, when b comes opposite to it, communicates with the ball d through the contact of k with f. in all other positions a, b, c and d are completely disconnected from each other. nicholson thus described the operation of his machine:-- "when the plates a and b are opposite each other, the two fixed plates a and c may be considered as one mass, and the revolving plate b, together with the ball d, will constitute another mass. all the experiments yet made concur to prove that these two masses will not possess the same electric state.... the redundant electricities in the masses under consideration will be unequally distributed; the plate a will have about ninety-nine parts, and the plate c one; and, for the same reason, the revolving plate b will have ninety-nine parts of the opposite electricity, and the ball d one. the rotation, by destroying the contacts, preserves this unequal distribution, and carries b from a to c at the same time that the tail k connects the ball with the plate c. in this situation, the electricity in b acts upon that in c, and produces the contrary state, by virtue of the communication between c and the ball; which last must therefore acquire an electricity of the same kind with that of the revolving plate. but the rotation again destroys the contact and restores b to its first situation opposite a. here, if we attend to the effect of the whole revolution, we shall find that the electric states of the respective masses have been greatly increased; for the ninety-nine parts in a and b remain, and the one part of electricity in c has been increased so as nearly to compensate ninety-nine parts of the opposite electricity in the revolving plate b, while the communication produced an opposite mutation in the electricity of the ball. a second rotation will, of course, produce a proportional augmentation of these increased quantities; and a continuance of turning will soon bring the intensities to their maximum, which is limited by an explosion between the plates" (_phil. trans._, 1788, p. 405). [illustration: fig. 4.--belli's doubler.] belli's doubler. nicholson described also another apparatus, the "spinning condenser," which worked on the same principle. bennet and nicholson were followed by t. cavallo, john read, bohnenberger, c.b. desormes and j.n.p. hachette and others in the invention of various forms of rotating doubler. a simple and typical form of doubler, devised in 1831 by g. belli (fig. 4), consisted of two curved metal plates between which revolved a pair of balls carried on an insulating stem. following the nomenclature usual in connexion with dynamos we may speak of the conductors which carry the initial charges as the field plates, and of the moving conductors on which are induced the charges which are subsequently added to those on the field plates, as the carriers. the wire which connects two armature plates for a moment is the neutralizing conductor. the two curved metal plates constitute the field plates and must have original charges imparted to them of opposite sign. the rotating balls are the carriers, and are connected together for a moment by a wire when in a position to be acted upon inductively by the field plates, thus acquiring charges of opposite sign. the moment after they are separated again. the rotation continuing the ball thus negatively charged is made to give up this charge to that negatively electrified field plate, and the ball positively charged its charge to the positively electrified field plate, by touching little contact springs. in this manner the field plates accumulate charges of opposite sign. [illustration: fig. 5.--varley's machine.] varley's machine. modern types of influence machine may be said to date from 1860 when c.f. varley patented a type of influence machine which has been the parent of numerous subsequent forms (_brit. pat. spec._ no. 206 of 1860). in it the field plates were sheets of tin-foil attached to a glass plate (fig. 5). in front of them a disk of ebonite or glass, having carriers of metal fixed to its edge, was rotated by a winch. in the course of their rotation two diametrically opposite carriers touched against the ends of a neutralizing conductor so as to form for a moment one conductor, and the moment afterwards these two carriers were insulated, one carrying away a positive charge and the other a negative. continuing their rotation, the positively charged carrier gave up its positive charge by touching a little knob attached to the positive field plate, and similarly for the negative charge carrier. in this way the charges on the field plates were continually replenished and reinforced. varley also constructed a multiple form of influence machine having six rotating disks, each having a number of carriers and rotating between field plates. with this apparatus he obtained sparks 6 in. long, the initial source of electrification being a single daniell cell. toepler machine. varley was followed by a.j.i. toepler, who in 1865 constructed an influence machine consisting of two disks fixed on the same shaft and rotating in the same direction. each disk carried two strips of tin-foil extending nearly over a semi-circle, and there were two field plates, one behind each disk; one of the plates was positively and the other negatively electrified. the carriers which were touched under the influence of the positive field plate passed on and gave up a portion of their negative charge to increase that of the negative field plate; in the same way the carriers which were touched under the influence of the negative field plate sent a part of their charge to augment that of the positive field plate. in this apparatus one of the charging rods communicated with one of the field plates, but the other with the neutralizing brush opposite to the other field plate. hence one of the field plates would always remain charged when a spark was taken at the transmitting terminals. [illustration: fig. 6.--holtz's machine.] holtz machine. between 1864 and 1880, w.t.b. holtz constructed and described a large number of influence machines which were for a long time considered the most advanced development of this type of electrostatic machine. in one form the holtz machine consisted of a glass disk mounted on a horizontal axis f (fig. 6) which could be made to rotate at a considerable speed by a multiplying gear, part of which is seen at x. close behind this disk was fixed another vertical disk of glass in which were cut two windows b, b. on the side of the fixed disk next the rotating disk were pasted two sectors of paper a, a, with short blunt points attached to them which projected out into the windows on the side away from the rotating disk. on the other side of the rotating disk were placed two metal combs c, c, which consisted of sharp points set in metal rods and were each connected to one of a pair of discharge balls e, d, the distance between which could be varied. to start the machine the balls were brought in contact, one of the paper armatures electrified, say, with positive electricity, and the disk set in motion. thereupon very shortly a hissing sound was heard and the machine became harder to turn as if the disk were moving through a resisting medium. after that the discharge balls might be separated a little and a continuous series of sparks or brush discharges would take place between them. if two leyden jars l, l were hung upon the conductors which supported the combs, with their outer coatings put in connexion with one another by m, a series of strong spark discharges passed between the discharge balls. the action of the machine is as follows: suppose one paper armature to be charged positively, it acts by induction on the right hand comb, causing negative electricity to issue from the comb points upon the glass revolving disk; at the same time the positive electricity passes through the closed discharge circuit to the left comb and issues from its teeth upon the part of the glass disk at the opposite end of the diameter. this positive electricity electrifies the left paper armature by induction, positive electricity issuing from the blunt point upon the side farthest from the rotating disk. the charges thus deposited on the glass disk are carried round so that the upper half is electrified negatively on both sides and the lower half positively on both sides, the sign of the electrification being reversed as the disk passes between the combs and the armature by discharges issuing from them respectively. if it were not for leakage in various ways, the electrification would go on everywhere increasing, but in practice a stationary state is soon attained. holtz's machine is very uncertain in its action in a moist climate, and has generally to be enclosed in a chamber in which the air is kept artificially dry. voss's machine. robert voss, a berlin instrument maker, in 1880 devised a form of machine in which he claimed that the principles of toepler and holtz were combined. on a rotating glass or ebonite disk were placed carriers of tin-foil or metal buttons against which neutralizing brushes touched. this armature plate revolved in front of a field plate carrying two pieces of tin-foil backed up by larger pieces of varnished paper. the studs on the armature plate were charged inductively by being connected for a moment by a neutralizing wire as they passed in front of the field plates, and then gave up their charges partly to renew the field charges and partly to collecting combs connected to discharge balls. in general design and construction, the manner of moving the rotating plate and in the use of the two leyden jars in connexion with the discharge balls, voss borrowed his ideas from holtz. wimshurst machine. all the above described machines, however, have been thrown into the shade by the invention of a greatly improved type of influence machine first constructed by james wimshurst about 1878. two glass disks are mounted on two shafts in such a manner that, by means of two belts and pulleys worked from a winch shaft, the disks can be rotated rapidly in opposite directions close to each other (fig. 7). these glass disks carry on them a certain number (not less than 16 or 20) tin-foil carriers which may or may not have brass buttons upon them. the glass plates are well varnished, and the carriers are placed on the outer sides of the two glass plates. as therefore the disks revolve, these carriers travel in opposite directions, coming at intervals in opposition to each other. each upright bearing carrying the shafts of the revolving disks also carries a neutralizing conductor or wire ending in a little brush of gilt thread. the neutralizing conductors for each disk are placed at right angles to each other. in addition there are collecting combs which occupy an intermediate position and have sharp points projecting inwards, and coming near to but not touching the carriers. these combs on opposite sides are connected respectively to the inner coatings of two leyden jars whose outer coatings are in connexion with one another. [illustration: fig. 7.--wimshurst's machine.] the operation of the machine is as follows: let us suppose that one of the studs on the back plate is positively electrified and one at the opposite end of a diameter is negatively electrified, and that at that moment two corresponding studs on the front plate passing opposite to these back studs are momentarily connected together by the neutralizing wire belonging to the front plate. the positive stud on the back plate will act inductively on the front stud and charge it negatively, and similarly for the other stud, and as the rotation continues these charged studs will pass round and give up most of their charge through the combs to the leyden jars. the moment, however, a pair of studs on the front plate are charged, they act as field plates to studs on the back plate which are passing at the moment, provided these last are connected by the back neutralizing wire. after a few revolutions of the disks half the studs on the front plate at any moment are charged negatively and half positively and the same on the back plate, the neutralizing wires forming the boundary between the positively and negatively charged studs. the diagram in fig. 8, taken by permission from s.p. thompson's paper (_loc. cit._), represents a view of the distribution of these charges on the front and back plates respectively. it will be seen that each stud is in turn both a field plate and a carrier having a charge induced on it, and then passing on in turn induces further charges on other studs. wimshurst constructed numerous very powerful machines of this type, some of them with multiple plates, which operate in almost any climate, and rarely fail to charge themselves and deliver a torrent of sparks between the discharge balls whenever the winch is turned. he also devised an alternating current electrical machine in which the discharge balls were alternately positive and negative. large wimshurst multiple plate influence machines are often used instead of induction coils for exciting rontgen ray tubes in medical work. they give very steady illumination on fluorescent screens. [illustration: fig. 8.--action of the wimshurst machine.] in 1900 it was found by f. tudsbury that if an influence machine is enclosed in a metallic chamber containing compressed air, or better, carbon dioxide, the insulating properties of compressed gases enable a greatly improved effect to be obtained owing to the diminution of the leakage across the plates and from the supports. hence sparks can be obtained of more than double the length at ordinary atmospheric pressure. in one case a machine with plates 8 in. in diameter which could give sparks 2.5 in. at ordinary pressure gave sparks of 5, 7, and 8 in. as the pressure was raised to 15, 30 and 45 lb. above the normal atmosphere. [illustration: fig. 9.--lord kelvin's replenisher. c, c, metal carriers fixed to ebonite cross-arm. f, f, brass field-plates or conductors. a, a, receiving springs. n, n, connecting springs or neutralizing brushes.] the action of lord kelvin's replenisher (fig. 9) used by him in connexion with his electrometers for maintaining their charge, closely resembles that of belli's doubler and will be understood from fig. 9. lord kelvin also devised an influence machine, commonly called a "mouse mill," for electrifying the ink in connexion with his siphon recorder. it was an electrostatic and electromagnetic machine combined, driven by an electric current and producing in turn electrostatic charges of electricity. in connexion with this subject mention must also be made of the water dropping influence machine of the same inventor.[1] the action and efficiency of influence machines have been investigated by f. rossetti, a. righi and f.w.g. kohlrausch. the electromotive force is practically constant no matter what the velocity of the disks, but according to some observers the internal resistance decreases as the velocity increases. kohlrausch, using a holtz machine with a plate 16 in. in diameter, found that the current given by it could only electrolyse acidulated water in 40 hours sufficient to liberate one cubic centimetre of mixed gases. e.e.n. mascart, a. roiti, and e. bouchotte have also examined the efficiency and current producing power of influence machines. bibliography.--in addition to s.p. thompson's valuable paper on influence machines (to which this article is much indebted) and other references given, see j. clerk maxwell, _treatise on electricity and magnetism_ (2nd ed., oxford, 1881), vol. i. p. 294; j.d. everett, _electricity_ (expansion of part iii. of deschanel's _natural philosophy_) (london, 1901), ch. iv. p. 20; a. winkelmann, _handbuch der physik_ (breslau, 1905), vol. iv. pp. 50-58 (contains a large number of references to original papers); j. gray, _electrical influence machines, their development and modern forms_ (london, 1903). (j. a. f.) footnote: [1] see lord kelvin, _reprint of papers on electrostatics and magnetism_ (1872); "electrophoric apparatus and illustrations of voltaic theory," p. 319; "on electric machines founded on induction and convection," p. 330; "the reciprocal electrophorus," p. 337. electric eel (_gymnotus electricus_), a member of the family of fishes known as _gymnotidae_. in spite of their external similarity the _gymnotidae_ have nothing to do with the eels (_anguilla_). they resemble the latter in the elongation of the body, the large number of vertebrae (240 in _gymnotus_), and the absence of pelvic fins; but they differ in all the more important characters of internal structure. they are in fact allied to the carps or _cyprinidae_ and the cat-fishes or _siluridae_. in common with these two families and the _characinidae_ of africa and south america, the _gymnotidae_ possess the peculiar structures called _ossicula auditus_ or weberian ossicles. these are a chain of small bones belonging to the first four vertebrae, which are much modified, and connecting the air-bladder with the auditory organs. such an agreement in the structure of so complicated and specialized an apparatus can only be the result of a community of descent of the families possessing it. accordingly these families are now placed together in a distinct sub-order, the ostariophysi. the _gymnotidae_ are strongly modified and degraded _characinidae_. in them the dorsal and caudal fins are very rudimentary or absent, and the anal is very long, extending from the anus, which is under the head or throat, to the end of the body. _gymnotus_ is the only genus of the family which possesses electric organs. these extend the whole length of the tail, which is four-fifths of the body. they are modifications of the lateral muscles and are supplied with numerous branches of the spinal nerves. they consist of longitudinal columns, each composed of an immense number of "electric plates." the posterior end of the organ is positive, the anterior negative, and the current passes from the tail to the head. the maximum shock is given when the head and tail of the _gymnotus_ are in contact with different points in the surface of some other animal. _gymnotus electricus_ attains a length of 3 ft. and the thickness of a man's thigh, and frequents the marshes of brazil and the guianas, where it is regarded with terror, owing to the formidable electrical apparatus with which it is provided. when this natural battery is discharged in a favourable position, it is sufficiently powerful to stun the largest animal; and according to a. von humboldt, it has been found necessary to change the line of certain roads passing through the pools frequented by the electric eels. these fish are eaten by the indians, who, before attempting to capture them, seek to exhaust their electrical power by driving horses into the ponds. by repeated discharges upon these they gradually expend this marvellous force; after which, being defenceless, they become timid, and approach the edge for shelter, when they fall an easy prey to the harpoon. it is only after long rest and abundance of food that the fish is able to resume the use of its subtle weapon. humboldt's description of this method of capturing the fish has not, however, been verified by recent travellers. electricity. this article is devoted to a general sketch of the history of the development of electrical knowledge on both the theoretical and the practical sides. the two great branches of electrical theory which concern the phenomena of electricity at rest, or "frictional" or "static" electricity, and of electricity in motion, or electric currents, are treated in two separate articles, electrostatics and electrokinetics. the phenomena attendant on the passage of electricity through solids, through liquids and through gases, are described in the article conduction, electric, and also electrolysis, and the propagation of electrical vibrations in electric waves. the interconnexion of magnetism (which has an article to itself) and electricity is discussed in electromagnetism, and these manifestations in nature in atmospheric electricity; aurora polaris and magnetism, terrestrial. the general principles of electrical engineering will be found in electricity