ACTION OF ABSORBENT MEDIA. 129 



492. Upon the hypothesis here assumed, the deposite of crystals becomes a phenom- 

 enon analogous to the curious configurations described by LICHTENBURG, when powders 

 are dusted on the surface of an electrified plate ; so close is the resemblance, that one 

 who sees crystallization produced by the sun for the first time, would be led, almost 

 involuntarily, to refer both to the same cause; suppose it granted, that when light falls on 

 any surface, that surface is electrified, it will exert an attraction on any particle in its vi- 

 cinity ; but if a conducting substance be placed in contact with the surface, not only will 

 it hinder deposite on the place which it occupies, but also it will rob the glass around 

 it for some distance ; here we find an explanation of the action of a tin foil ring. 

 Again, if that conducting substance be so placed as to cast its shadow on the glass, no 

 deposite should take place on that shadow, nor for a certain distance around it, because 

 the electricity of the adjacent parts would pass towards the unelectrified spaces, thus 

 conferring, by a surface conduction, a low charge to all the shaded parts. 



493. We meet, however, if we pass beyond these simple explanations, with so many 

 difficulties, that we are not encouraged to seek farther confirmation of this hypothesis ; 

 there are some facts which prove, almost demonstratively, that electricity is not the agent 

 in question. If, instead of a ring of rosin, we make use of a ring of sealing-wax or a 

 ring of pitch, these, though they are non-conductors, do not fail to protect ; the action 

 of a metallic ring, when placed inside of a jar, cannot, so far as I know, receive any ex- 

 planation, especially if we are to admit the non-conducting power of a space filled with 

 camphor vapour only. It is plain and obvious that transparency and opacity have no- 

 thing to do with it ; glass and rosin, it is true, do not protect ; but oil, which is equally 

 transparent, protects as powerfully as a metal. 



494. Are we to refer this singular action to the rays of light, to the rays of heat, or 

 to the chemical rays ! By the action of absorbent media, attempts have been made to 

 determine this question. A barometer tube (fde,jig. 91) had a conical tube fixed on 

 its outside, so that the interstice could contain liquids at c d, without leaking. Into this 

 torricellian vacuum I passed a piece of camphor, and exposed the arrangement to the 

 sun ; having filled the interstice with water, it was found to have crystals on the aphe- 

 lion side, there being a ring of them, as at e e, jig. 92, all round the tube. The water 

 was then poured out, and a solution of sulphate of copper and ammonia introduced. On 

 examination, it was found that on the side nearest the sun no crystals were to be seen, 

 but on the other side there was a dense layer of them, extending exactly half way round 

 the tube, and very much resembling the shape of Jig. 93. A yellow liquid, the bichro- 

 mate of potassa, was next introduced : a result to all appearance exactly like the former 

 was again produced; but having observed that the thickness of the media had a very 

 sensible effect, apparently due to their becoming warm, and not casting off their caloric 

 with sufficient rapidity by radiation, I made an alteration in the arrangement, by inter- 

 posing between the torricellian vacuum and the light a trough capable of containing the 

 different solutions. This trough being filled with solution of bichromate of potassa, and 

 the ray tested that it could not blacken chloride of silver, in about one hour the tube 

 presented the following appearance : there were some pretty large crystals which ex- 

 tended round the tube, as at a fig. 94, which, on the aphelion side, suddenly mounted 



R 



