SCIENTIFIC WRITINGS. 397 



In a later essay on the dependence of the electric 

 conductivity of carbon on temperature I have confirmed 

 Matthiessen's assertion, that the conductivity of carbon 

 increases with rising temperature, and have shown the 

 objections of Beetz and Auerbach to be erroneous. 

 In explanation of this surprising behaviour of carbon 

 I advanced the hypothesis that the different states of 

 carbon charcoal, graphite, diamond - - are allotropic 

 states of "carbon devoid of latent heat" not occurring 

 in Nature, and are essentially distinguished from one 

 another by the qitantity of absorbed latent heat. 



This hypothesis was further confirmed and deve- 

 loped by an investigation of the property of selenium, 

 discovered by Willoughby Smith, of being a better 

 conductor of electricity in the light than in the dark. 

 I found that besides the selenium, which is changed 

 by a slight enhancement of temperature from the 

 amorphous non-conducting into the crystalline con- 

 ducting condition, there is still a third modification, 

 which is produced by heating amorphous selenium a 

 long time till near its melting point, i. e. to about 

 400 F. Both these modifications of the electricity- 

 conducting selenium are essentially distinguished from 

 one another by this, that the former conducts electro- 

 lylically, i. e. like the electrolytic fluid conductors, 

 better at a higher temperature, the second, long and 

 highly heated, on the other hand metallically, i. e. 

 like the metals worse at a higher temperature. In 

 this behaviour of amorphous selenium, rapidly cooled 

 from the fused condition viz. when heated to over 



