182 
IOWA ACADEMY OP SCIENCE 
agency. After the third filing on the sample just mentioned, the conductivity 
of the sample was still changing slowly after a period of more than two months. 
It is not so clear however whether all the change is of the nature of recovery 
toward the same condition prevailing before the filing process began. There 
was probably 0.06 gram of selenium on the sample of the cell. After removing 
somewhat more than one-third of this amount and waiting for equilibrium the 
conductivity was 0.14, or only about one-twentieth of what it was before the 
removal. It would seem from this that a large part of the change was not a 
recovery. Another set of obesrvations for another sample of selenium is shown 
by the curves in Fig. 2. By the third filing which took place on the fifth day 
less than one-half of the selenium was removed and yet the conductivity was 
reduced from 7 to 1.7 finally. It may be that not only the removed selenium 
but also that near the surface is rendered incapable of conducting. As is well 
known* the light-sensitive selenium is reduced by powdering to the non-con- 
ducting amorphous form. The surface of the cell does present a black, glaring 
appearance, like the vitreous selenium, after the filing. However it seemed to 
me that the surface color changed during the course of a month to the usual 
gray. 
THE NATURE OF THE EFFECT COMPARED TO OTHER EFFECTS. 
The only other known mechanical effects that might be related to the one 
described in this paper, are probably the pressure effect and the radium effect. 
Pressure does decrease the resistance to a most remarkable extent** but the effect 
displays little or no hysteresis. The resistance returns to its original value 
very quickly. In fact I know of no other effect from which the recovery is so 
rapid. However Monten later observed that with pressures as large as 30,000 
atmospheres there was a slight hysteresis effect. This difference between the 
pressure effect and the rupture effect is probably what is implied in the names. 
The hydraulic pressures applied were uniform on the surface and so took no 
part of the selenium beyond the elastic limit, while by filing or sand-blasting 
parts of the selenium were actually broken off, and perhaps a large part of the 
surface crystals were strained beyond the elastic limit. 
One might expect that the rays from radium would produce an effect of the 
nature of rupture by abrasion. In the paper by Brown and Stebbins the (loc. cit) 
the results of short exposures to radium of 2,000,000 activity are described. 
The' decrease of resistance was of the same order of magnitude as that produced 
by abrasion. Also the recovery is slower than it is from light exposures, but 
not so slow as the recovery from filing. Therefore it is quite possible that the 
radium effect is a rupture effect. A study of long exposure of selenium to 
radium, perhaps many days, would give further evidence on this point. No 
doubt the resistance would Continue to decrease until the radium were removed,, 
instead of reaching an equilibrium condition in a few minutes as happens with 
light exposures of corresponding intensities. 
*See paper by A. P. Saunders, Jour. Phys. Chem. 4, p. 423. 
**See paper by Brown ' and ■ Stebbins, Phys. Rev. 26, p. 273, and also by Monten» 
Archiv for Math., Astr., och Fysik, 4, p. 1, 1908. 
