688 



Mr. W. H. Jenkinson on 



charge, are given off and in the absence of any conductivity 

 in the air a large P.D. between the electrodes would result 

 from this, the unshielded one being positive. The production 

 of this in a vacuum is well known. In the present case, 

 however, the P.D. will be largely annulled hy the conductivity 

 imparted simultaneously to the air by the X rays 5 but it was 

 still necessary to see whether it affected the curves to any 

 extent. To test this, instead o£ wires, two copper plates 

 5 cm. long and 1 cm. broad were used, first placed trans- 

 versely to the beam of X rays, when the effect would be a 

 maximum, and then turned through a right angle so that 

 they presented only a minimum surface to the beam. No 

 systematic difference between the two curves was found, 

 consequently this effect, and also any due to secondary 

 X rays, which might affect the ionization in the immediate 

 neighbourhood of the electrodes, must be negligibly small. 



With X rays considerable difficulties due to the impossibility 

 of keeping the tube constant over the long time required to 

 plot a complete curve were met with, and the experiments 

 were finally continued with radium. A 10 mgm. tube of the 

 bromide was placed 4*5 cm. from the electrodes, which were 

 then shielded in turn as already described. The results of a 

 number of experiments in which different thicknesses of Al 

 were used for shielding are given in Table I. Each number 

 in column 2 is the mean value of the deflexion obtained on 

 reversal taken from a curve of 4 or 5 reversals as shown in 

 fig. 2. In the 3rd column the corresponding observed E.M.F. 

 of the cell, Vac , is given in volts. 



Table I. 





2 Voo 







! 



X 



Thickness 

 of Al in cms. 



Electrometer 



scale-divisions 



on reversal. 



m Voo 



in volts. 



a 

 K 



Em 



volts. 











•005 



5-7 



•0041 



•0364 



•0047 



•010 



11-7 



•0085 



•0317 



•0098 



•015 



14-2 



•0103 



•0378 



•0116 



•020 



16-5 



•0120 



•0342 



•0137 



•0325 



17-5 



•0127 



•0369 . 



•0144 



•0525 



24-7 



•0179 



•0308 



•0208 



•065 



29-2 



•0212 



•0280 



•0250 



•147 



44-75 



•0325 



•0303 



•0378 



•27 



51-5 



•0372 



•0265 



•0440 



•472 



53-45 



•0388 



•0247 



•0465 



•674 



56-06 



•0407 



•0250 



•0487 



1096 



58-75 



•0427 



•0245 



•0512 



Sensitiveness of elctrometer 1 scale-division = '001455 volt. 



