410 //. .1. flu instead — Emission of Electrons by 



ments with the two metals on the ring, F. Fig. 4 gives a 

 graphical representation of these results; it is to be remem- 

 bered, however, that, the ordinates of the different curves are 

 not exactly adjusted to each other, on account of the differ- 

 ences just mentioned. It is indeed impossible to make satis- 



Fig. 4. 



CM. 

 4 



• AIR 



ALUMINIUM 



x qoo 



!l) 



20 



40 



60 



80 



100 



ttO 



140 



160 



factory correction for these differences without knowing the 

 relative values of the " incidence" and " emergence" S-radia- 

 tion ; for, between the generation of the two, the a-rays pass 

 through the leaf on the electrode — in the one case aluminium, 

 in the other case, gold. It is plain, however, that this correc- 

 tion (if it could be made) would bring closer together the two 

 curves in their decreasing portions, while it would not much 

 affect their increasing portions where they are already in close 

 agreement. 



The results of Taylor, mentioned at the beginning of this 

 paper, would lead us to expect that the metal curves would lie 

 to the left of the air-curve and have a less pronounced knee, 

 as appears to be the case. But they give equal grounds for 

 anticipating that the gold curve should be to the left of that 

 obtained with aluminium, and this is not confirmed by the 

 experiments. Quite apart from these somewhat hypothetical 

 considerations, it does not appear probable that two metals 

 which differ so much in atomic weight and in other properties 

 should give effects so nearly identical ; the ionization curves in 

 different gases are markedly different even when the gases 

 differ much less from each other than gold and aluminium. 



