230 Glow-Lamps raised to High Incandescence. [Mar. 26, 



of playing about the heel, surrounds apparently the whole filament. 

 The result is that the current passes through the galvanometer and 

 through the rarefied space. This is clearly shown in Experiment 1. 



14. It is quite clear that the critical point is reached when the 

 filament commences to be disintegrated by the projection of its 

 molecules from its surface. It is here that the resistance of the fila- 

 ment commences to increase, and the law of radiation and light 

 emission (" Proc. Roy. Soc," No. 229, 1884) commences to be 

 departed from, as was shown by me in a paper read before the British 

 Association at the Montreal meeting.* 



15. It is very evident that this Edison effect is due to the formation 

 of an arc between the carbon filament and the metal plate fixed in the 

 vacuous bulb ; that this arc is due to the projection of the carbon 

 particles in right lines across the vacuous space ; and that it makes its 

 appearance earlier, and is more strongly marked, when the connex- 

 ions are as shown in Fig. 1 than when they are reversed, because, as 

 Mr. Crookes has pointed out, the projection proceeds from the nega- 

 tive to the positive pole, and it would commence at the point of least 

 resistance. Its presence is detrimental to the life of the lamp, and as 

 its appearance is contemporaneous with the blue effect, the latter is a 

 warning of the approach of the critical point, and a sure indication 

 that the electromotive force is dangerously high. It is also clear that 

 as the Edison effect is only evident when we are " among the breakers," 

 it is not available for practically regulating the conditions of electric 

 light currents as its ingenious discoverer originally proposed. 



* In this paper I pointed out from experimental data that the light emitted by 

 a glow-lamp varied apparently as the sixth power of the current. I verified this 

 law, not only by subsequent experiments of my own, but, which is much more 

 satisfactory, by experiments of others. Professor Kittler, of Darmstadt, and Captain 

 Abney made, independently of each other, most careful and exhaustive measure- 

 ments in this direction. I tabulated and traced them out in curves. They fully 

 confirm the law that 



L = &C 6 , 



but within limits, and that these limits embrace the ordinary range of a glow-lamp 

 when used for artificial illumination. As long as the resistance and the current vary 

 uniformly together the law holds good ; but as the state of incandescence is 

 increased, a point is reached, Tarying with each kind of lamp, when the resistance 

 ceases to diminish at the same rate, and eventually increases. When this occurs 

 the law is departed from, and the light emitted increases less than the sixth 

 power of the current. The filament speedily breaks. The point of departure from 

 the law indicates a point when a change of state occurs in the carbon filament. 

 Disintegration probably sets in. This point ought to be determined for each kind of 

 lamp, and it should never be allowed to be reached, for it is from this point that 

 decay commences and rupture follows. 



The Society then adjourned over the Easter Recess to Thursday, 

 April 16th. 



