﻿584 Mr. Karl T. Compton on Contact Difference of 



of 80 volts to the field between and B shifts the curve to 

 the left about one volt. This indicates that the perforated 

 foil is not a perfect shield, and that a part of the field 

 existing on the upper side of C makes itself felt below C *. 

 If this " stray field " explanation is correct, one would 

 expect that if the perforations in C were larger, the amount 

 of stray field existing on its under side would be increased. 

 This was tested by obtaining a series of curves similar to 

 those in fig. 4 (PI. IX.), but using the platinum wire gauze 

 box instead of the perforated foil box for C. In this second 

 case the proportion of open to closed space in the box was 

 decidedly larger than in the first case. Now it was' found 

 that the curves were shifted about 1*5 volts to the left for 

 each 80 volts added between C and B, as against one volt to 

 80 in the first case. This confirms the explanation given. 



Since, then, the field between C and B exerts some small 

 effect on the under side of the perforated box C the curves 

 which are given in the following sections are all shifted 

 somewhat to the left of their true positions. Curves (/i) 

 and (i) and the method of finding the true positions of the 

 curves will be explained later (section 6). It may be noted 

 here that the curves to the right in fig. 4 (PI. IX.) are more 

 erect than are those to the left. This is explained simply 

 and consistently with the view we have taken, when one 

 remembers that the electrons which are moving most rapidly 

 when they reach C are the most likely to strike and stick to 

 C itself. It takes a stronger field to bend their paths so as 

 to draw them through the perforations than to do this for 

 the slow moving electrons. Thus, in the case of the curves 

 to the left, practically all the electrons are drawn through 

 the perforations in C, while farther to the right the curves 

 indicate that a larger proportion of the high velocity electrons 

 fail to get through. Hence the curves to the left should be 

 less erect and should more nearly represent the true shape 

 of the " distribution of velocity" curve. That this is true 

 receives additional confirmation in section 7. 



From these results it was evident that to compare the 

 curves obtained when different metallic boxes were used 



* Tins is apparently in contradiction to the results found by E. Laden- 

 burg and K. Markau {Phijs. Zeit. ix. p. 821, 1908), who found, with an 

 arrangement somewhat similar to this, that, as the difference of potential 

 above the perforated screen was increased, the curves shifted to the left 

 until the difference of potential was 50 volts. After this they found the 

 position of the curves to be independent of the difference of potential 

 above the screen. They explained this result as being due entirely to 

 the prevention of electron reflexion and not to the stray field through 

 the perforations. 



