232 



NA TURE 



[July 3, 1890 



carbon molecule or small carbon mass just detached from 

 an incandescent surface behaves in the same way and 

 has a greater grip for negative than for positive charge ? 

 If this can possibly be assumed, w^e can complete our 

 hypothesis as follows: — Consider a carbon molecule or 

 small collection of molecules just set free by the high 

 temperature from the negative leg of the incandescent 

 carbon horseshoe. This small carbon mass finds itself in 

 the electrostatic field between the branches of the incan- 

 descent carbon conductor (see Fig. 14). It is acted upon 

 inductively, and if it behaves like the hot iron ball in 

 Prof. Guthrie's experiment it loses its positive charge. 

 The molecule then being charged negatively is repelled 

 along the lines of electric force against the positive leg. 



n\ 



Fig. 14. — Rough diagram illustrating a theory of the manner in which 

 projected carhon molecules may acquire a negative charge. 



The forces moving it are electric forces, and the repetition 

 of this action would cause a torrent of negatively-charged 

 molecules to pour across from the negative to the positive 

 side of the carbon horseshoe. If we place a metal plate 

 in their path, which is in conducting connection with the 

 positive electrode of the lamp carbon, the negatively 

 charged molecules will discharge themselves against it. 

 A plate so placed may catch more or less of this stream 

 of charged molecules which pour across between the heels 

 of the carbon loop. There are many extraordinary facts, 

 which as yet I have been able only imperfectly to explore, 

 which relate to the sudden changes in the direction of the 

 principal stream of these charged molecules, and to their 



Fig. 15. — Electric arc projected by magnet against a third carbon, and 

 showing a strong electric current flowing through a galvanometer, g, 

 connected between the positive and this third carbon. 



guidance under the influence of magnetic forces. The 

 above rough sketch of a theory must be taken for no 

 more than it is worth, viz. as a working hypothesis to 

 suggest further experiments. 



These experiments with incandescence lamps have pre- 

 pared the way for me to exhibit to you some curious facts 

 with respect to the electric arc, and which are analogous 

 to those which we have passed in review. If a good 

 electric arc is formed in the usual way, and if a third in- 

 sulated carbon held at right angles to the other two is 

 placed so that its tip just dips into the arc (see Fig. 15), 

 we can show a similar series of experiments. It is rather 

 more under control if we cause the arc to be projected 



NO. 1079, VOL. 42] 



against the third carbon by means of a magnet. I have 

 now formed on the screen an image of the carbon poles 

 and the arc between them, in the usual way. Placing a 

 magnet at the back of the arc, I cause the flame of the 

 arc to be deflected laterally and to blow against a third 

 insulated carbon held in it. There are three insulated 

 wires attached respectively to the positive and to the 

 negative carbons of the arc, and to the third or insulated 

 carbon, the end of which dips into the flame of the arc 

 projected by the magnet. On starting the arc this third 

 carbon is instantly brought down to the same electrical 

 potential as the negative carbon of the arc, and if I con- 

 nect this galvanometer in between the negative carbon 

 and the third or insulated carbon, I get, as you see, no 

 indication of a current. Let me, however, change the 

 connections and insert the circuit of my galvanometer in 

 between the positive carbon of the arc and the middle 

 carbon, and we find evidence, by the violent impulse given 

 to the galvanometer, that there is a strong current flowing 

 through it. The direction of this current is equivalent to 

 a flow of negative electricity from the middle carbon 

 through the galvanometer to the positive carbon of the 

 arc. We have here, then, the " Edison effect " repeated 

 with the electric arc. So strong is the current flowing in 

 a circuit connecting the middle carbon with the positive 

 carbon that I can, as you see, ring an electric bell and 

 light a small incandescent lamp when these electric-current 



Fig. 16. — Galvanometer g and battery B inserted in series between negative 

 carbon of electric arc and a third carbon to show unilateral conductivity 

 of the arc between the negative and third carbons. 



detectors are placed in connection with the positive and 

 middle carbons. 



We also find that the flame-like projection of the arc 

 between the negative carbon possesses a unilateral con- 

 ductivity. I join this sm.all secondary battery of fifteen 

 cells in series with the galvanometer, and connect the two 

 between the middle carbon and the negative carbon of 

 the arc. Just as in the analogous experiment with the 

 incandescent lamp, we find we can send negative elec- 

 tricity along the flame of the arc one way but not the 

 other. The secondary battery causes the galvanometer 

 to indicate a current flowing through it when its negative 

 pole is in connection with the negative carbon of the arc 

 (see Fig. 16), but not when its positive pole is in connec- 

 tion with the negative carbon. On examining the third 

 or middle carbon after it has been employed in this way 

 for some time, we find that its extremity is cratered out 

 and converted into graphite, just as if it had been em- 

 ployed as the positive carbon in forming an electric arc. 



Time forbids me to indulge in any but the briefest re- 

 marks on these experiments ; but one suggestion may be 

 made, and that is that they seem to indicate that the 

 chief movement of carbon molecules in the electric arc is 

 fro7n the negative to the positive carbon. The idea sug- 

 gests itself that, after all, the cratering out of the positive 

 carbon of the arc may be due to a sand-blast-like action 



