8o 



NATURE 



[May 2 2, 1890 



Upon the whole, however, we have ended by reaching a much 

 more satisfactory state of agreement than seemed possible when 

 we began. For Prof. Lankester now says he deems it " certain 

 that some cases must sometimes occur in which the selection- 

 mean is larger than the birth-mean," and that as regards such 

 cases I have his " full concurrence in stating that the cessation 

 of selection leads to dwindling." And as he previously agreed 

 that cessation of selection leads also to a loss of shape and disin- 

 tegration of structure, the only question that remains between us 

 is as to whether there are any cases in which completely developed 

 organs cease to present variations of size below the standard of 

 full efficiency, and therefore will remain unaffected by the with- 

 drawal of the selection by which they were evolved. But this 

 is a question which does not vitally affect the principle of pan- 

 mixia ; and it only remains to add that I do fully "reciprocate " 

 what he has said as to there being " no ill-feeling between us." 

 George J. Romanes. 



Photo-electric Impulsion Cells. 



Before publishing in detail the results of many experiments on 

 the generation of electricity by the action of light falling on 

 certain sensitive substances, I wish to make known a result 

 which seems to be of a most remarkable character. 



In this communication I shall give merely enough information 

 to enable a reader to understand the special result which I 

 desire now to make known. 



The photo-electric cell which I employ consists of a small 

 glass tube, represented in the figure, filled with an alcohol ; 

 two metallic plates, P and Q, are immersed in the liquid ; each 



plate is connected with a platinum wire which may either be 

 soldered to the plate or passed through a small hole in the plate 

 axiA. pinched tightly to it ; these wires pass through the ends of 

 the glass tube and are sealed into it. The poles of the cell are 

 A, B, and these are connected with the poles of a quadrant 

 electrometer (Clifton's form of Thomson's). 



The plate p is sensitized by a peculiar process, the mere 

 publication of the details of which would not enable a reader to 

 make it successfully. The publication of the process is therefore 

 reserved for a future occasion. The plate Q is quite clean — not 

 sensitized to light. The cell is fixed vertically in a clamp (not 

 represented in the figure). "When the cell is of the " impulsion " 

 kind, what happens is as follows. Daylight (represented by the 

 arrow l) being allowed to fall on the sensitive plate P, the spot 

 on the scale of the electrometer moves, and after a few seconds 

 comes to rest, indicating an electromotive force varying with the 

 intensity of the light, its amount for such diffused daylight as 

 we have at present (May 10) at noon being between J a volt 

 and I of a volt — which is, I submit, a surprisingly great 

 magnitude. On the withdrawal of the light, the deflection falls, 

 and there are means of rapidly getting rid of the deflection 

 without injury to the cell. Either before or after this deflection 

 caused by light ceases, let a slight tap (sometimes inaudible) be 

 given to the base or clamp in which the cell rests, and then 

 results a remarkable change in the cell. It is no longer sensitive to 

 light. This insensitive state is indicated by a rapid return motion 

 of the spot on the scale ; it is merely indicated by this motion, 

 there being no necessary connection between this motion and 

 the insensitive state, for if the cell were now left for some time 

 NO. 1073, VOL. 42] 



(perhaps an hour or so) in the dark, the disturbing E.M.F. of 

 the cell would vanish, and there would be nothing to tell us 

 that the cell remains insensitive ; but that it is really still in the 

 insensitive state we find at once on again exposing it to light. 

 Another gentle tap given to the clamp, or the stone table on 

 which the whole apparatus rests, will restore the sensitive state ;. 

 and so on indefinitely, the sensitive and insensitive states 

 following each other and being produced, in the case of many 

 such cells, with great ease. 



These results I found a long time ago, and they have been 

 seen by or communicated to several scientific friends. From 

 the first, I maintained that the results are due to an alteration of 

 the molecular state of the sensitive surface, or of the layer of 

 contact of this surface with the liquid, and that in one arrange- 

 ment of the molecules the light energy can be taken up 

 electrically, while it cannot be so taken up in the other. In 

 my first experiments the plates were tightly pinched to the 

 platinum wires— not soldered, as soldering endangered the 

 sensitive layer — and the obvious objection was made that 

 "loose contacts" were unsatisfactory. I have several results, 

 however, which dispose of this objection even in the case of very 

 loose contacts ; but I may set the matter at rest by saying that 

 I have been able to make soldered junctions, and with them to 

 obtain the results. 



I now come to the special point which is the occasion of this 

 communication. A few days ago I was investigating the effect 

 of static charges communicated to the plates on the sensitive and 

 insensitive states, and in the course of these experiments I found 

 that if a Voss machine, not in any way connected with the cell 

 or the electrometer, was worked in the room while the cell was 

 in the insensitive state, the moment a spark 

 passed between the poles of the Voss, the insen- 

 sitive state was altered to the sensitive, whether 

 the cell was connected with the electrometer or 

 not. Finally, I found that the best method of 

 showing the inductive effect of the spark is to 

 connect an insulated wire, w, apparently of any 

 length, to either pole (A in the figure) of the 

 cell, and to place the poles, g, h, of the Voss 

 near the wire (a distance of several feet will do 

 with a spark about half an inch long). If g and 

 h are two or three feet from any part of the 

 wire w, a spark about one-eighth of an inch long 

 suffices to change the cell from the insensitive 

 to the sensitive state. 



The effect is not one on the electrometer, nor 

 is it due to sound, and I have repeated the 

 results with several cells many scores of times 

 before people interested in them. At present I 

 am endeavouring to produce by electro-magnetic 

 induction the reverse change, viz. that from the 

 sensitive to the insensitive state ; but, although 

 such must apparently be possible, I have not yet succeeded. 



The sudden alteration of the insensitive to the sensitive state 

 is produced in a most marked manner by the spark of a Hertz 

 oscillator at as great a distance as the laboratory room in which 

 I work allows. This distance is usually only about eight or 

 ten feet, but I observed the change effected occasionally when 

 the oscillator was at a distance of some thirty feet or more. In 

 this latter case, however, the action was interfered vyith by the 

 unavoidable presence of wires along the walls, &c., intervening 

 between the Hertz and my impulsion cell. 



If the cause to which I have assigned the change from the 

 photo-electrically insensitive to the photo-electrically sensitive 

 state of the cell is the true one, it is impossible to avoid the 

 speculation that impulsion results of this kind may be very 

 common in the economy of Nature ; and that the mode in which 

 solar energy is taken up by plants may be affected, and even 

 altered in kind, by sudden electro-magnetic disturbances. The 

 effect of a Hertz oscillation is, indeed, not confined to an 

 alteration of a plate from the insensitive to the sensitive state ; 

 for I have cells in which if the sensitive plate is, on exposure 

 to light, electrically negative to the back plate, a Hertz oscillator 

 at a distance will reverse the relation when the plate is again 

 exposed to light. George M. Minchin. 



Royal Indian Engineering College, Cooper's Hill, 

 May 10. 



p.S. — While the above communication was going through the 

 press, I made an experiment which renders it almost certain that 

 in the impulsion cells the results are due to the formation of some 



