July 3, 1890] 



NATURE 



2x1 



and the middle plate. In that circuit the cell cannot send 

 any sensible current at all, as it is at the present moment 

 connected up. But if we reverse the direction of the cell 

 so that its positive pole is in connection with the middle 

 plate, the galvanometer at once gives indications of a 

 very sensible current. This highly vacuous space, lying 

 between the middle metal plate on the one hand, and the 

 incandescent carbon on the other, possesses a kind of 

 unilateral conductivity, in that it will allow the current 

 from a single galvanic cell to pass one way but not the 

 other. It is a very old and familiar fact that in order to 

 send a current from a battery through a highly rarefied 

 gas by means of metal electrodes, the electromotive force 

 of the battery must exceed a certain value. Here, how- 

 ever, we have indication that if the negative electrode by 

 which that current seeks to enter the vacuous space is 

 made incandescent the current will pass at a very much 

 lower electromotive force than if the electrode is not so 

 heated. 



A little consideration of the foregoing experiments led 

 to the conclusion that in the original experiment, as de- 

 vised by Mr. Edison, if we could by any means render 

 the middle plate very hot, we should get a current flowing 

 through a galvanometer when it is connected between the 

 middle plate and the negative electrode of the carbon. 

 This experiment can be tried in the manner now to be 

 shown. Here is a bulb (Fig. 12) having in it two carbon 



resistance of the rest of the circuit is not large. We may 

 embrace the foregoing facts by saying that if the elec- 

 trodes, but especially the negative electrode, which form 

 the means of ingress and egress of a current into a vacu- 

 ous space are capable of being rendered highly incan- 

 descent, and if at that high temperature they are made to 

 differ in electrical potential by the application of a very 

 small electromotive force, we may get under these cir- 

 cumstances a very sensible current through the rarefied 

 gas. If the electrodes are cold a very much higher 

 electromotive force will be necessary to begin the dis- 

 charge or current through the space. These facts have 

 been made the subject of elaborate investigation bv 

 Hittorf and Goldstein, and more recently by Elster and 

 Geitel. It is to Hittorf that I believe we are indebted 

 for the discovery of the fact that by heating the negative 

 electrode we greatly reduce the apparent resistance of a 

 vacuum. 

 ! Permit me now to pave the way by some other experi- 

 j ments for a little more detailed outline of the manner in 

 which I shall venture to suggest these negative molecular 

 I charges are bestowed. This is really the important 

 i matter to examine. In seeking for some probable ex- 

 1 planation of the manner in which these wandering mole- 

 cules of carbon in the glow-lamp bulb obtain their negative 

 charges, I fall back for assistance upon some facts dis- 

 covered by the late Prof. Guthrie. He showed some 

 years ago new experiments on the relative powers of 

 incandescent bodies for retaining positive and negative 

 charges. One of the facts he brought forward 1 was that 



Fig. 12.— Experiment showing that when the " middle plate " is a carbon 

 loop rendered incandescent by insulated battery b, a current of negative 

 electricity flows from .M to the positive leg of main carbon c across the 

 vacuum. 



loops ; one of these is of ordinary size, and will be 

 rendered incandescent by the current from the mains. 

 The other loop is very small, and will be heated by a 

 well-insulated secondary battery. This smaller incan- 

 descent loop shall be employed just as if it were a middle 

 metal plate. It is, in fact, simply an incandescent middle 

 conductor. On repeating the typical experiment with this 

 arrangement, we find that the galvanometer indicates a 

 current when connected between the middle loop and 

 either the positive or the negative terminal of the main 

 carbon. I have little doubt but that if we could render 

 the platinum plate in our first-used lamp incandescent by 

 concentrating on it from outside a powerful beam of 

 radiant heat we should get the same result. 



A similar set of results can be arrived at by experi- 

 ments with a bulb constructed Hke an ordinary vacuum 

 tube, and having small carbon loops at each end instead 

 of the usual platinum or aluminium wires. Such a tube 

 is now before you (see Fig. 13), and will not allow the 

 current from a few cells of a secondary battery to pass 

 through it when the carbon loops are cold. If, however, 

 by means of well-insulated secondary batteries we render 

 both of the carbon loop electrodes highly incandescent, a 

 single cell of a battery is sufficient to pass a very con- 

 siderable current across that vacuous space, provided the 

 NO, 1079, VOL. 42] 



Fig. 13. — Vacuum tube having carbon loop electrodes, c c, at each end 

 rendered incandescent by insulated batteries, b, b^, showing current 

 from Clark cell, Ck, passing through the high vacuum when the 

 electrodes are incandescent. 



a bright red-hot iron ball, well insulated, could be charged 

 negatively, but could not retain for an instant a positive 

 charge. He showed this fact in a way which it is very 

 easy to repeat as a lecture experiment. Here is a gold- 

 leaf electroscope, to which we will impart a positive 

 charge of electricity, and project the image of its divergent 

 leaves on the screen. A poker, the tip of which has been 

 made brightly red-hot, is placed so that its incandescent 

 end is about an inch from the knob of the electroscope. 

 No discharge takes place. Discharging the electroscope 

 with my finger, I give it a small charge of negative elec- 

 tricity, and replace the poker in the same position. The 

 gold leaves instantly collapse. Bear in mind that the ex- 

 tremity of the poker, when brought in contiguity to the 

 knob of the charged electroscope, becomes charged by 

 induction with a charge of the opposite sign to that of the 

 charge of the electroscope, and you will at once see that 

 this experiment confirms Prof. Guthrie's statement, for 

 the negatively charged electroscope induces a positive 

 charge on the incandescent iron, and this charge cannot 

 be retained. If the induced charge on the poker is a 

 negative charge, it is retained, and hence the positively 

 charged electroscope is not discharged, but the negatively 

 charged electroscope at once loses its charge. Pass in 

 imagination from iron balls to carbon molecules. We 

 may ask whether it is a legitimate assumption to suppose 

 the same fact to hold good for them, and that a hot 



' " On a New Relation between Electricity ar.d Heat," P/ii7. Ma£., vol. 

 xlv. p, 308 (1873). 



