328 SECTIONAL TRANSACTIONS.— A. 



Dr. J. A. V. Fairbrother. — A new method for investigating conduction 

 phenomena in semi-conductors (12.5). 



Using a tungsten wire coated with a layer of insulating material, a method 

 is described whereby Langmuir probe methods of measurement in a low- 

 pressure mercury discharge can be made to yield information on the 

 electrical behaviour of the insulating material. 



For pure aluminium oxide the electrical conductivity is found to vary 

 with temperature in accordance with A. H. Wilson's theory of electronic 

 semi-conductors. The Wilson energy difference W 2 — Wj is found to be 

 2-9 electron volts between 1280 K. and 1725 K. and 6-6 electron volts at 

 higher temperatures. The critical temperature 1725 K. is approximately 

 the sintering point of alumina. 



Experimental evidence is given in support of the discovery of electron 

 diffusion through alumina. Mention is made of the bearing of this pheno- 

 menon on the experiments of Reimann, Treloar and E. F. Lowry, relating 

 to the thermionic emission from oxide coated cathodes. 



The conduction currents flowing across an alumina layer 0-005 in. 

 thick between tungsten and plasma in the forward and reverse direction 

 are given at different temperatures for voltages up to 250 volts. At high 

 temperatures the conductivity is greater in the direction of electron flow 

 from metal to oxide to plasma. 



Afternoon. 



Excursion to works of Metropolitan Vickers Electrical Co., Ltd., 

 Trafford Park, Manchester. 



Tuesday, September 15. 



Dr. L. Vegard. — Changes of intensity distribution within the auroral 

 spectrum due to sunlight and other causes (10.0). 



The spectrum of the auroral luminescence is mainly composed of the 

 following parts : — 



(a) The green auroral line (\ = 5577). 



(b) The red lines 6300, 6365, of which the first predominates. 



(c) The first negative group of nitrogen. 



(d) The first positive group of nitrogen. 



(e) The second positive group of nitrogen. 



The observed effects of intensity variations may be conveniently classified 

 as follows : — 



(1) Change of intensity distribution with altitude. In 1923 the writer 

 found an increase in the ratio c to a with increase of altitude. 



(2) Type effects. It has, e.g., been found that the ratios a/d and a/c 

 are smaller for certain diffuse areas than for the ordinary distinct 

 radiant forms. 



(3) Intensity variations for the same type — usually accompanied by 

 colour changes. The red aurora of type A was found to be due to 

 an enormous enhancement of the line b relative to the other parts, 

 and type B was due to enhancement of red bands of the first positive 

 group. 



(4) Recently it has been found by the writer in collaboration with E. 

 T^nsberg that in a sunlit atmosphere the auroral spectrum shows 



