544 Mr. H. P. Walmsley on the Distribution of the 



anode activity. If between two plane parallel surfaces n 

 deposit atoms per second per unit volume are produced, then 

 in all £nH atoms per unit area are directly projected on to 

 the anode by recoil *, where It is the range o£ recoil of the 

 projected product. Assuming this formula for a cylindrical 

 surface, we get 



Anode activity ^ira . \n& _ 3*04 



Cathode activity ~~ ir(b 2 -a 2 )~7p ~ TOO' 



where a= radius of inner electrode, b= radius of case, and 

 p= density of active matter in the gas. Inserting the 

 following values, 



R — O'l mm., Z> = 6*75 mm., a = l*61 mm., 

 we get n/p = 16 approximately. 



Thus, if the anode activity were due to a surface condensa- 

 tion of emanation, the concentration must be equivalent to a 

 layer of emanation-bearing gas y 1 ^ mm. thick of at least 16 

 times the concentration in the rest of the volume. It seems 

 very unlikely that such a state would occur with widely 

 differing metals as lead, aluminium, tin, and brass. It also 

 follows that if this concentration does not exist, the amount 

 of active deposit projected on to the anode is of the order of 

 Y 1 ^ of the observed anode activity, i. e., about 0*2 per cent, 

 of the cathode activity. This is the value obtained under 

 similar circumstances in the case of actinium t, which seems 

 to justify the use of the above formula for purposes of 

 approximation. 



Too much weight must not be placed on the values given 

 in the table, as the measurements of the anode activity 

 depend upon the state of the surfaces. A brass rod was 

 carefully cleaned with sand-paper and two electrodes were 

 -cut from it. One was subsequently highly polished with 

 rouge, and the two were placed in the chambers. The rods 

 were earthed and the cases charged to —960 volts. It was 

 found that the polished rod constantly gave an a-ray 

 activity about 5 per cent, greater than the others, although 

 both rods must have collected the same amount of active 

 matter. As there is no means of obtaining, exactly the 

 same degree of polish on two rods, or of estimating the 

 difference in this quality, it is difficult to say to what extent 

 the values given above are affected. The rods in all cases 

 were well polished, so perhaps the effect of polish on the 

 results is small. 



* Rutherford, 'Radioactive Substances/ p. 403. 

 t Walmsley, Phil. Mag. xxvi. p. 381 (1913). 



