Resistance of Selenium with Temperature. 469 



modification depending on the time during which it is 

 maintained at one temperature. So that with a change in 

 temperature there is a change in the quantity of the various 

 modifications, the equilibrium stage being arrived at very 

 slowly *. After this, if the temperature is allowed to remain 

 constant, the degree of stability of each variety increases 

 or the velocity of further transformation decreases, and it 

 becomes more and more sluggish to change that condition. 

 That this is perhaps the actual case, may be realized from 

 the fact that a cell, maintained for a long time at one tem- 

 perature and subsequently subjected to a small change of 

 temperature, requires a very long time to assume its ultimate 

 conductivity appropriate to that temperature — much longer, 

 for example, than in the experiments to which fig. 1 refers. 



5. Observations with Selenium Crystals. 



The conclusions drawn from the above experiments being- 

 well supported by the modified theory of transformation, and 

 it being well known that the process of annealing is nothing 

 but a crystallization of the amorphous variety, it was hoped 

 to repeat the experiments with crystals produced by the con- 

 densation of molten selenium at various temperatures. The 

 method suggested by Brown f was adopted, but as the 

 crystals so formed were not large enough to test them indi- 

 vidually, no quantitative experiment was pursued. Of those 

 mentioned by Brown, three different varieties of crystals 

 were recognized : — 



(1) Needle-shaped crystals. 



(2) Groups of white acicular crystals. 



(3) Groups of reddish acicular crystals. 



Of these, the first variety showed a striking behaviour. 

 Several of these formed a network and were deposited on 

 the wall of the tube. They appeared red in the reflected 

 light and greyish black in the transmitted light, and had a 

 moderate conductivity at room temperature. On slightly 



* The velocity of such transformations is very slow, particularly at 

 low temperature. But Berger (Zeitschrift fur Anorganische Chemie, 

 vol. lxxxv. p. 75 (191.4), has shown that by the presence of certain 

 catalytic agents as silver selenide, the rate of transformation is greatly 

 increased. This therefore indirectly supports the theory of trans- 

 formation. The influence of the catalytic agent in increasing the 

 velocity of transformation may eventually explain the more marked 

 effect near the electrodes, as observed by S. Bidwell (Chemical News. 

 vol. lii. p. 191) and more recently by White (Phil. Mag' (6) vol. xxvii. 

 p. 370). 



t Brown, Phys. Rev. vol. iv. p. 8o (1914), 



