34 On the Electrical Resistance of Selenium Cells. 



above referred to) did not at ordinary temperatures exhibit 

 this peculiarity. When heated, its resistance at once went 

 down without any preliminary rise. This cell was placed in 

 air at a temperature of 0°, and after remaining for half an 

 hour its resistance was found to be 147,000 ohms. The tem- 

 perature was then slowly raised ; and, as I expected, the resist- 

 ance at first went up, attaining a maximum of 219,000 ohms 

 at 13°, after which it went down to 134,000 ohms at 36°, when 

 the experiment was stopped. The curve fig. 2, which is on 

 the same scale as the others, shows the results in a very stri- 

 king manner, altogether excluding the possibility of accidental 

 disturbance. This particular cell differed from others only in 

 the fact that it acquired its maximum resistance at a tempe- 

 rature slightly below instead of slightly above the average 

 temperature of the air. 



The supposition that light produces its effect by heating is 

 further negatived by the fact, that a comparatively high 

 degree of temperature is required to bring down the resistance 

 of the cell to the point to which it is instantly reduced by ex- 

 posure to a strong light. When a selenium cell is for a 

 moment exposed to sunlight, it does not become perceptibly 

 warm to the touch ; but the amount of dark heat necessary to 

 effect the same reduction in its resistance as is caused by 

 a moment's sunshine would certainly render it too hot to 

 handle. 



Again, those who have experimented with the photophone 

 know well that the best results are obtained only when pre- 

 cautions are taken to exclude those rays which are especially 

 instrumental in producing heat, as by filtering the beam of 

 light through a solution of alum. Dark radiation does indeed 

 per se diminish the resistance of selenium ; but the diminu- 

 tion due to dark radiation is to some extent masked by the 

 rise of temperature which accompanies it, and which gene- 

 rally tends to produce the opposite effect. 



To me it seems clear that the electrical effects of radiation 

 are, in this case at least, no more due to the intermediate 

 action of temperature than are the chemical effects which 

 radiation sometimes produces, as in the various photographic 

 processes. All such effects are no doubt ultimately of a mecha- 

 nical nature; but while increased temperature may result from 

 vibrations the periodicity of which vary between very wide 

 limits, the other effects arise only when there is some more 

 or less definite relation between the period of the aether- waves 

 and the molecular constitution of the substance upon which 

 they act. 



