196 



KNOWLEDGE & SCIENTIFIC NEWS. 



[AuGi'ST, 1905. 



The Irvduction Pump. 



WitK Suggestions as to Reversal 

 in Influence Machines. 



Bv Charles E. Benham. 



In two previous articles published in " Knowledge" 

 (November, 1904, and May, 1905) some simple experi- 

 ments in electric induction were described, the ijeneral 

 principle of which was that of the original " doubler" 

 of Abraham Bennet, and also to some extent that of 

 the majority of devices which succeeded the doubler, 

 with a view to render the multiplied-charge process 

 available in a practical way. Such devices include the 

 machines invented by X'arley, Toepler, Voss, and others, 

 as well as such modifications as Lord Kelvin's water 

 dropping apparatus, and finally the admirable arrange 

 ment adopted by Mr. Wimshurst, and known by his 

 name. 



The multiplying process, by which these machines 

 acquire such powerful electric charges, is briefly attri- 

 butable to the fact that while an insulated charged body 

 can confer, by induction, without losing its own elec- 

 tricity, successive charges of opposite sign upon two 

 or more conductors successively earthed when near it, 

 it can obviously be made to receive in return an increase 

 of its own charge if it in turn is placed within the in- 

 ductive sphere of those two or more charged bodies so 

 that they all re-act upon it simultaneously. This is the 

 principle that underlies the action of most of the 

 influence machines, the difference being chiefly in the 

 way in which this cycle of action and re-action is 

 brought about. 



It was shown in the experiments already referred to 

 that there are alternative ways of carrying out the 

 process. I*'or instance, the simultaneous re-action of 

 two or more inductively charged bodies may be pro- 

 vided for cither by bringing them into actual contact 



with each other by super-position, or by merely placing 

 the conductor that has to receive their conjoint influence 

 in an intermediate position so as to be within the sphere 

 of all of tliem. The latter process lends itself more 

 readily to practical application. For example, if .i 

 series of strips of tinfoil are attached to the under side 

 of a piece of glass, as in Fig i, A, B, C, D, E, they may 

 each be given an induced charge by the single piece of 

 insulated tinfoil F, on the upper side of a similar glass, 

 if this plate is drawn successively over A, B, C, D, 

 and E, earthing each as the front edge of F passes over 

 it. Then, placing the plate bearing F over the other, 

 and touching F, it receives from A. B, C. D, and E a 

 combined influence, increasing its original charge. On 

 repeating this cycle a few times the tinfoils, which 

 had no measurable initial charge, become strongly elec- 

 trified, the sign of F's electricity being the opposite of 

 that of A, B, C, D, and E. Reducing the experiment to 

 a still simpler form, the tinfoils A. B, C. D, and E may 

 Ix? dispensed with, and the finger mav be placed on the 

 under side of the plain glass instead, drawing it along 

 as F is moved, so that it successively occupies the posi- 



tions of A, \i, C, D, and E (see Fig 2), and it will do 

 duty for the tinfoils. The charge is retained by the 

 glass surface, and after a few strokes from end 

 to end, alternated with earthings of the upper tinfoil, 

 Ihe charges will accumulate. There is no need to move 

 the upper plate at all during the charging process. Let 

 it rest on the plain glass, and draw the outstretched 

 finger lightly over the under surface of this several 

 limes, each time touching the upper tinfoil after the 

 operation, and a considerable charge will accumulate 

 on the upper plate and will be given off by its tinfoil 

 when the two plates are parted and a conductor is pre- 

 sented to the upper one. Tliat the charge is not due to 

 friction from drawing the finger over the lower glass i.^ 

 evident from the fact that sometimes the charges will 

 be positive and sometimes negative on the upper plate. 

 This variation of the polarity is very curious, and it 

 is difficult to associate it with a definite cause. Two 

 sucli pairs of induction plates may be made exactly alike, 

 and kept near each other under precisely similar con- 

 ditions. On testing their action at different times the 

 upper plates of the two pairs will sometimes be found 

 oppositely chaigcd and sometimes similarly. A numbc 

 of observations made at varir)us limes of the day, and 

 under various atmospheric conditions, failed to show 

 any agreement of behaviour on the part of the two pairs 

 of plates. .Somct rnics they would retain the same 

 polarity for several days. .Sf)mctimes one would change 

 c\ery few hours and the other would continue constant 

 for a long w hilc. At other times each would vary at very 

 short intervals, and f)ccasionally one would be so in- 

 ..onstant that it was difiTicult to make it acquire an accu 

 mulated charge. In 70 observations taken at variouj 

 times of the day during the space of a week, the upper 



