222 Prof. Minchin's Experiments in Photoelectricity. 



the experiments closely satisfy an equation of this form, which 

 is, of course, that of a logarithmic curve. 



The curve of fall is probably of the same nature, but the 

 rate of fall is much slower than that of rise, as is evident by 

 an inspection of the diagram exhibited to the Society. 



Unhomogeneous Surfaces. — To produce a homogeneous sur- 

 face on the sensitive plate, it is manifestly necessary to satisfy 

 two conditions : the deposit produced by the oxidizing solu- 

 tion must be uniform, and so must the heating by the spirit- 

 lamp. It is obvious, then, that if these conditions are not 

 fulfilled, we must be prepared to find some part of the plate 

 positive and another part negative on exposure to light ; and 

 such has actually been found to be the case in many experi- 

 ments. But a much more extraordinary result of quite 

 common occurrence is that described under the next head. 



Impulsion- Cells. — It often happens that, a few days after a 

 cell is mounted and found to be duly sensitive to light, an 

 exposure produces no E M.F. at all. But if a slight tap 

 (sometimes scarcely audible) is given either to the support of 

 the cell or to the table on which this support rests, a change, 

 indicated by the motion of the spot on the electrometer- scale, 

 takes place in the cell, and it is as sensitive to light as it was 

 originally. Another tap given to the base throws the cell 

 again into the insensitive state ; another tap will restore the 

 sensitive state ; and so on indefinitely. 



These results are not due to any defect in the contact of the 

 platinum wires with the plates ; these wires are tightly 

 pinched to clean parts of the plates, and often soldered to 

 tbem. 



It might be conjectured that these results are due to the 

 formation of some gas in the cell ; but bow can one tap 

 replace the gas which has been displaced by another tap ? 



1 believe the sensitive and insensitive states to be due to 

 some molecular alteration, either in the sensitive surface or 

 in the liquid, or in their layer of contact — a sort of polarity 

 in the medium which, as in the case of magnetic bodies, can 

 be produced or destroyed by vibrations. Thus the E.M.F. 

 due to light may be simply due to strain and not accompanied 

 by any actual chemical combination. 



A most remarkable instance of these impulsion-effects may 

 be cited to show that they are not due to any defective 

 contact. Referring to fig. 3, which represents the sensitive 

 plate completely submerged in the liquid, let the plate be only 

 partially submerged, and let the liquid surface be at a point 

 C, about midway between B and A, the portion CA of the cell 

 being occupied by vapour alone. This will be the state of 



