1890.] between Electrodes at different Temperatures. 



119 



the carbon conductor, a sensitive galvanometer connected between 

 this insulated metal plate enclosed in the vacuum and the external 

 positive electrode of the lamp indicates a current of some milli- 

 amperes passing through it when the lamp is set in action, bat the 

 same instrument when connected between the negative electrode of 

 the lamp and the insulated metal plate indicates no sensible current. 

 This phenomenon in carbon incandescence lamps was first observed by 

 Mr. Edison, in 1884, and further examined by Mr. W. H. Preece, in 

 1885.* The primary object of the experiments described in this 

 paper was the further examination of this effect, but the inquiry 

 has extended itself beyond this range and embraced some general 

 phenomena of electric discharge between electrodes at unequal tem- 

 peratures, and in particular has revealed some curious effects in the 

 behaviour of an electric arc taken between carbon poles towards a 

 third insulated carbon on metal poles. 



The first series of experiments had reference to the nature of the 

 effect observed in the incandescence lamps having an insulated wire 

 or plate placed in the vacuum. 



If a platinum wire is sealed through the glass bulb of an ordinary 

 carbon filament lamp and carries at its extremity a metal plate, so 

 placed as to stand np between the legs of the carbon horseshoe with- 

 out touching either of them, then when the lamp is actuated by a 

 continuous current it is found that r — 



(1.) This insulated metal plate is brought down instantly to 

 the potential of the base of the negative leg of the carbon, and 

 no sensible potential difference exists between the insulated metal plate 

 and the negative electrode of the lamps, whether the test be made by 

 a galvanometer, by an electrostatic voltmeter, or by a condenser. 



(2.) The potential difference of the plate and the positive electrode 

 of the lamp is exactly the same as the working potential difference of 

 the lamp electrodes, provided this is measured electrostatically, i.e.} 

 by a condenser, or by an electrostatic voltmeter taking no current, 



* See « Koy. Soc. Proe.,' vol. 38, 1885, p. 219. " On a Peculiar Behaviour of Glow 

 Lamps when raised to High incandescence." — In this paper Mr. Preece describes a 

 very careful series of observations carried out with Edison incandescence lamps, 

 and which cover tbe same ground as a portion of the experiments here described. 

 The results given in (4), (7), and (11) confirm the facts which were first ascertained 

 by him. He also arrived at the general conclusion that the phenomena so observed 

 are due to an electric convection by matter projected from the incandescent carbon. 

 By carrying up the working electromotive force of the lamp to a point productive of 

 very high incandescence, he was able to measure the resulting current through a 

 galvanometer connected between the positive lamp electrode and the middle plate 

 corresponding to every degree of incandescence, and showed that, whilst increasing 

 up to a certain point, the galvanometer current fell off rapidly soon after a certain 

 critical temperature was reached, which corresponded to the appearance of a blue 

 light or haze in the glass receiver. — [Jan. 14th, 1890.} 



