SURFACE PROPERTIES OF GERMANIUM 17 



gas ambient and therefore no dependence on the corresponding changes 

 in surface dipole. This result was somewhat unexpected and the first 

 time the experiment was performed it was hard to believe that the 

 previously measured changes in c.p. actually were taking place. In this 

 case the gas ambient was experimented with to try to change Vs and it 

 was found that Vs could be changed from the order of 10^ cm/sec to 

 greater than 10^ cm/sec and back again by exposure of the filament to 

 (NH)40H fumes and then HCl fumes respectively.^ While interesting, 

 this has no direct bearing on the other experiments. The experiment 

 was performed again with freshly prepared rods and this time the 

 cycling used in the c.p. experiments was rigidly adhered to, giving Vg 

 equal to 70 cm/sec and 200 cm/sec approximately for samples A and 

 D respectively. The experiments were then repeated again using new 

 filaments cut from the samples as close as possible to the surfaces used 

 in the c.p. experiments leading to the results shown in Fig. 11, namely, 

 Va equal to 50 cm/sec and 170 cm/sec respectively. From these experi- 

 ments it was concluded that Va is approximately constant in the range 

 involved and is determined by the nature of the sample and the surface 

 treatment used. It was noted in some of these experiments that Va for 

 the first cycle was somewhat larger than for the subsequent cycles. 

 This change when it occurred is probably to be correlated with the 

 changes in the early cycles in the c.p. measurements. Similar measure- 

 ments on sample C gave Vg equal to 1500 cm/sec. No measurements of 

 Va were made on samples B and E. 



V. Other experimental measurements 



The specific resistance of each sample was measured near the surface 

 used in the experiments. It w^as approximately constant across the 

 surface but did vary slowly with depth in some of the samples. 



The body life times were measured on each sample. The thickness of 

 the slices used was intentionally made large compared to their cor- 

 responding diffusion lengths, about 0.5 cm for A, B, C and E and about 

 2.0 cm for D. The mobilities were taken from J. R. Haynes^ measure- 

 ments: Mn = 3600 and Mp = 1700 cmVvolt sec. There is some uncer- 

 tainty as regards. the exact value of the equilibrium product of holes 

 and electrons, np, at 300°K. We have used the value 6.3 X 10^® obtained 

 from some unpublished data of G. L. Pearson. 



The light source was calibrated by replacing the germanium sample 

 with one of F. S. Goucher's n-p junctions.^ The bell jar, with every- 

 thing else including the reference electrode, was left in their normal 



