SOI// (O.V// l//M/v'.//x') .;/>/ /.V(/s /V /7/ls/(s (/ 117 



till- .iikhIi- ami llir citlioiU- an- two par.illi'l platr^. </ cciiliini'irc^ 

 apart, and tluTt- an- P hfliiim atoms in a rul)ic ci-nlinu-tri' of tlic Ras 

 iK'twit-n. and Sn i'lirtrons start out in a sirond dirrrtly lowartls the 

 an(KU' from any area of the ratli<Hie, llie proportion AA' .Vn of elec- 

 tron^ whirl) .\rc inti'rce|>led before they reach ilic anode is 



A.V .Vo=l 



(1) 



and the niiinlier of electrons reachini; the corresponding; area on the 

 anode in a second, .V»— A.V, conforms to the e(iu,itioii: 



log, (A'o-A.Y)= -.4 A/ + const. 



(2) 



The coefficient .1 is a constant to be interpreted as the effective cross- 

 sectional area of the helium atom relatively to an oncoming electron — 

 that is, the atom behaves towarils the electron like an obstacle pre- 

 senting the impenetrable area .-1 to it. 



In the experiments perfortned to verify these assertions and de- 

 termine the \alue of .1, the simple geometricifl arrangemeni whicli 

 I have described is generally modified in one way or another for 

 greater accurac>- or convenience. Mayer approached most nearly 

 to the sim()le arrangement; in his apparatus (Fig. 1) the electrons 



Fig. 1 — Apparatus for determining the percentage of electrons which go across a 

 gas of variable thickness without interception. (Mayer, AnnaJen der Physik) 



which e.iierge from the hot lllamenl al G, pass througli the two slits 

 in front of it, and then go down the long tube to the anode A', which 

 is drawn backward step by step. The logarithinic curves of current 

 versus distance for various pressures of nitrogen (Fig. 2) are straight. 

 Unfortunately the current also diminishes as the distance is increased 

 when the nitrogen is pumped out altogether: this is attributed partly 



