FIXING ELECTRICAL FIGURES. 405 



paper ; the effect was the same, but it seemed to me more 

 feeble. 



8. A solution of nitrate of silver was poured over the sur- 

 face of an electrised plate, so as to form a bath on its surface ; 

 a rake, formed of ten common pins, was made to touch the 

 glass with its points along the course of the invisible image ; 

 the silver was of course precipitated in an arborescent form, 

 and I thought it probable that the lines of deposition might 

 follow the course of the invisible image, but I could not be 

 certain of any such effect, though in one experiment there 

 seemed to be some slight indication of it. 



The above experiments were repeated many times, both 

 with positive and negative electricity from the coil. At first 

 I believed I had found a remarkable difference in the effect of 

 the change of direction of the discharge in the cases where 

 hydrofluoric acid was employed ; and in two experiments, in 

 which the tinfoil of the upper glass plate was connected with 

 the positive terminal, the impression of the word on the lower 

 side of the upper plate after exposure to the vapour was 

 polished, while the residue was frosted ; and in two subsequent 

 cases, the upper tinfoil being connected with the negative 

 terminal, the reverse was the case. In subsequent experi- 

 ments, however, great irregularities took place in this relation, 

 and it seemed to depend on the time of exposure and on 

 slight differences in the distance between the paper letters 

 and the glass, the latter not being ^brought into perfectly 

 uniform contact with the surfaces of the glass. 



After the first few experiments I placed a marble paper- 

 weight on the upper glass, and found the effects more uniform 

 and perfect. 



An electrisation for periods of from five to ten minutes 

 produced the sharpest and clearest effects ; when the electri- 

 sation was prolonged, a blur or second margin gradually 

 appeared, and increased in extent around the outline of the 

 letters, having somewhat the appearance which would have 

 been presented had the paper letters been moist, and the 

 liquid slightly extended itself from their edges over the glass. 



When the electrisation was thus prolonged the figures 

 were visible on inclining the glass to the light, without 



