1886.] Conduction and Molecular Composition, fyc. 281 



is the best conductor at ordinary temperatures. On reference to the 

 curves, it will be seen that the resistance of both silver chloride and 

 bromide suddenly increases when the change from the fused to the 

 solid state sets in ; but that no such change takes place in the case of 

 the iodide. Silver iodide fuses at 527° according to Rodwell, but at 

 about 540° according to Kohlrausch ; its electrical resistance increases 

 only gradually after it has become solid, and remains almost a linear 

 function of the temperature during an interval of 400°, until suddenly 

 at near 150° it increases enormously, this change taking place at the 

 moment when according to Rodwell (" Phil. Trans.," 1882, p. 1133) 

 it passes from the transparent, plastic, amorphous solid to the opaque, 

 brittle, crystalline state, the volume increasing considerably as shown 

 by the annexed curve. Kohlrausch has proved most conclusively that 



the solid iodide may undergo electrolysis. It would seem that almost 

 immediately after solidification in the case of silver chloride and 

 bromide practically the whole mass consists of complex aggregates so 

 constituted as to be exceedingly bad conductors, but that such aggre- 

 gates are formed much less readily by silver iodide. 



Metallic Conduction. 



I do not propose, in any way to discuss metallic conduction $ but 

 merely to call attention to some of the analogies between it and 

 electrolytic conduction. 



It is conceivable, and it would appear probable from the fairly 

 regular manner in which the electrical resistance of most pure metals 



