HOLE CONCENTRATION AND POINT CONTACTS 483 



electrodes at the two ends so that the tield pulling the holes along the 

 filament could be varied. The concentration of holes was determined from 

 the change in resistivity of that segment of the filament between the two 

 probes. Measurements of admittance were made by passing a small current 

 between the two probes connected in series. The area of the filament is 

 about 1.6 X 10~^ cm- and the normal resistance between the probes 

 about 1800 ohms. The normal conductivity is thus 



ffo = .009/(1800 X 1.6 X 10-^) = 0.03 (ohm cm)-\ (39) 



As shown in Fig. 6, Pearson finds a linear relation between G and pn- 

 The line best fitting Pearson's data is 



G = Go + (8 X 10-«) {pa/m) (mhos). (40) 



The theoretical value of the coefficient may be obtained from Eq. (38). 

 Taking 



a = 1, iS = 0.5, ao = 0.03 



h = 2.0, A = 10-« cm2, r = 5 X lO"" cm, (41) 



we get 



a^aoA/b r^ = 15 X 10-^ mhos. (42) 



Pearson's data, represented by (40), apply to the conductance of two 

 point contacts in series, and the conductance of each one may be about 

 twice that given by (40). Thus the theoretical value is in good agreement 

 with the observed. There is no indication that a differs from unity at low 

 voltage. 



Suhl varied the concentration of holes in the vicinity of probe points by 

 application of a transverse magnetic field as well as by injection from an 

 emitter point. The experiment is illustrated in Fig. 7. He used a filament 

 with a cross-section of about .025 X .025 cm. Four probe points were 

 placed along the length of the filament at intervals of about .04 cm. A 

 total current of 4 ma flowed in the filament. 



In one experiment, none of this current was injected, so that the con- 

 centration of holes was normal in the absence of the magnetic field. 

 Measurements were made of the floating potentials and of the conduct- 

 ances of the probe points. Then a transverse magnetic field was applied 

 and the conductances measured again. We are interested here only in the 

 case of a large field (30,000 gauss) in such a direction as to sweep the 

 holes to the opposite side of the filament. Suhl believes that under these 

 conditions the concentration of holes near the probe points is practically 

 zero. The difference between the conductances with and without the field 



