August, 1905.] 



KNOWLEDGE & SCIENTIFIC NEWS. 



197 



plates of the two pairs of inductors g:ave the following 

 results : — 



First pair. Second pair. 



43 times neg'ative. 53 times negative. 



26 times positive. 17 times positive. 



I time variable. 



The induced charge of the upper plate was thus more 



frequently negative than positive in both cases. If the 



charge had been due to friction by drawing the finger 



along the under plate, that plate, being well coated with 



shellac, would have been charged negatively, and the 



induced electricitv' in the upper plate would have been 



positive, showing again that the charge could not have 



been initiated by friction. 



As is well known, the Wimshurst machine, or any 

 other self-exciting influence machine, is in the same way 

 variable as to the polarity it will assume at starting 

 after it has been left at rest for some time. 



Returning now to the experiments described in the 

 two former articles, the object of the present contribu- 

 tion is to explain how the same process (briefly recapi- 

 tulated above) mav be utilised more effectively for 



Fig. 3- 



pumping up from the earth really practical supplies of 

 electricity; how, in fact, a machine may be made on 

 this principle, and incidentally how such a machine as 

 the Wimshurst is a contrivance of this very character. 

 In the Wimshurst the disposition of the brushes and 

 oppositely rotating discs gives, as was shown in the 

 first article, a double multiplying arrangement of char- 

 acter similar to that which we have performed by hand 

 more tediously in the experiments that have been 

 described in this and the preceding articles. It follows, 

 indeed, that two oppositely rotating discs, with 

 sectors, as in the \Mmshurst, ought, theoretically, to 

 become charged with electricity of opposite sign with 

 only a single brush to each disc placed as at A and B 



.\ moment's consideration of this diagram will show 

 that the directions of rotation of the respective discs 

 being as shown by the arrows, the sectors successivel;, 

 charged by induction at the brush A will (two or three 

 of them at least) re-act simultaneously on the other plate 

 at B, when they have travelled to the position in front 

 of that brush. Inversely the sectors charged succes- 

 sively at B will several of them act simultaneously upon 



the sector at \, when they have reached that spot. .4t 

 A and B, therefore, there will be two points at which 

 continually increasing induction charges of opposite 

 sign will be received. It will be found, indeed, that, 

 arranging the Wimshurst apparatus in this way, 

 with only two brushes, theory- is exactly borne 

 out by experience. Tlie brushes promptly glow, 

 and the two discs are oppositely charged. it 

 is this latter fact, however, that renders their 

 charge unavailable under such conditions. The 

 opposite charges of the two discs hold each other 

 ''bound," and consequently the collectors are not able 

 to draw off any charge from the plates when arranged 

 in this way with the two single brushes. The four 

 brushes of the Wimshurst provide for opposite charges 

 on different parts of each disc, with a consequence that 

 at certain parts of the revolution (where the collecting 

 combs are placed) the adjacent part of each disc is 

 charged with electricity of the same sign, and this, of 

 course, is not " bound," but is readily taken off by the 

 collectors. It is in this way that the Wimshurst 

 machine is so efficiently adapted to its purpose, the 

 charges being " bound " at such portions of the revolu- 

 tion as is appropriate, and set free only at such portions 

 as are necessary for collection of the charge. 



It is, however, obvious that if it were possible to 



Fig. 4. 



liberate the "bound" charges on the two discs 

 arranged with the two single brushes as shown in Fig 

 3, we should have an efficient accumulator capable o'^ 

 charging jars, as in the case of the Wimshurst. 



This liberation can be effected by providing that a 

 portion of each disc shall be removed from its com- 

 panion's influence. If, for example, the discs are 

 placed as in Fig 4, the induction can still be effected 

 by placing brushes at A and B, where the discs over- 

 lap, while the collecting can be accomplished at the free 

 parts of the discs, viz., at C and D. It will be necessarv', 

 however, in order that the sectors may travel past each 

 other in opposite directions, that the discs themselves 

 should both rotate the same way, as shown by the 

 arrows in the diagram. 



This device is a true induction pump, drawing, in 

 chain pump fashion, from the earth contrary electrici- 

 ties in each of the respective discs with its circle of 

 sectors. 



Two discs of 16 inches diameter arranged in this 

 way, with 24 sectors on each, will give a torrent of 

 5-inch sparks, and the machine is readily self-exciting. 



A troublesome, and perhaps unexpected, difficulty, 



