464 BELL SYSTEM TECHNICAL JOURNAL 



More than this, some of the very data presented by Compton and 

 Langmuir support the generaHzed checkerboard theory as we have 

 presented it, i.e., taking into account that both the black and the 

 white squares may be covered with thorium but to different extents. 

 On page 155, Compton and Langmuir discuss two log i vs. V F curves 

 obtained by Reynolds^" for thoriated tungsten. Curve A in their 

 Fig. 4 is a "normal" curve while curve C is taken after the surface has 

 been bombarded by positive ions. They state "this bombardment 

 must have roughened the surface and there is evidence that it also 

 fractured the surface layer of tungsten crystals." In our notation this 

 means that h has been decreased because of the roughening and m 

 has been increased because the amount of thorium removed in some 

 spots was larger than that removed in others. Now the decrease in h 

 should shift the region at which the curve approaches the Schottky 

 line to higher values of 7*" or V F\ while the increase in ix should result 

 in a steeper curve and should decrease log iao- (See our Fig. 16.) 

 But this is precisely what curve C does. 



Reynolds^" in discussing this same data says: "The effect of bom- 

 bardment was a semi-permanent one. Subsequent activation and 

 deactivation by temperature (below 2700°) shifted the curve along the 

 current axis but did not otherwise alter its unique character. Flashing 

 at 2700° K. or higher, where rapid sintering of tungsten is known to 

 take place, destroyed the effect of bombardment and subsequent 

 activation produced normal log i vs. F* curves." Every detail of this 

 behavior is just what is to be expected on our view; the "semi-perma- 

 nent" efifect is caused by the decrease in h which does not become 

 normal until the damage to the crystals has been repaired by high 

 temperature treatment; the shifting of the curves along the current 

 axis is caused by changes in p and /x brought about by activation and 

 deactivation. 



On page 156 Compton and Langmuir^ discuss Kingdon and Lang- 

 muir's data for thoriated tungsten at various degrees of activation 

 {d or /) and at various temperatures. Some of the results are shown 

 in their Fig. 5 which is a plot of T log i vs. V F. They point out three 

 and only three distinctive features of these curves. All three support 

 the checkerboard theory. They say: "At the highest field-strengths 

 (about 10,000 volts /cm.) the curves are seen to approach the theoretical 

 slope." Our analysis shows that this means that all surfaces have 

 about the same h irrespective of Q and T. This is predicted by our 

 theory since b is determined by the crystal size which is independent 

 of/ and T. In discussing curv-es for a constant/ they sa>': " In every 

 case the departures from the Schottky line become greater as the 



