^Resistance of 



Measured. 



current =i 

 micros. 



meg. 



b + V 



f direct 

 ( inverted 



•0945 

 •0942 



c + c* 



f direct 



•1756 



\ inverted 



•1746 



s + s' 





•5028 



with Crystalline Selenium. 407 



Current =0'42 

 microw. 



meg. 



•1022 



•1021 



•1777 



•1775 



•5032 



The agreement is in favour of the selenium, the mean resist- 

 ance of which does not appear to change by decreasing the 

 current-strength : the mean resistances of the junctions, how- 

 ever, increased, one of them considerably ; and it is therefore 

 probable that in them, and not in the selenium, lies the change 

 in question. 



Resistance altering ivith strength of Current. — Professor W. 

 G. Adams has pointed out that when the battery-power is in- 

 creased the apparent resistance of the selenium is diminished. 

 In some of the specimens of selenium which I tested, I found 

 that while the current was weak, up to a certain limit, the 

 resistance increased in one direction and decreased in the other 

 direction ; but after passing the limit in question, the resist- 

 ance decreased in both directions for any further increase of 

 current. 



One plate of selenium, annealed at 200° C, was kept at a 

 constant temperature and measured in a Wheatstone bridge, 

 the proportional resistances of which were respectively 1000 

 and 100,000 ohms. The selenium was inverted each time by 

 a commutator, so that measurements were repeated in both 

 directions, whilst the currents in the bridge-resistances always 

 went in the same direction. 



Current in 

 selenium. 



Measured resistance. 



Current direct. 



Current inverted. 



microwebers. 



2 



4 



6 



8 



10 

 12 

 14 

 20 

 30 



megohm. 

 •4107 

 •4119 

 •4128 

 •4131 

 •4133 

 •4133 

 •4133 

 •4130 

 •4126 



megohm. 

 •4093 

 •4080 

 •4072 

 •4063 

 •4056 

 •4050 

 •4047 

 •4038 

 •4023 



A similar result was found with a second plate, the resist- 



