TEANSACTIONS OF SECTION B. 591 



result that the colour of the cloth on each side of the pieces of iron was completely 

 discharged, and the colour so abstracted was arranged in deeply coloured lines, run- 

 ning in a semicircle from each end of the iron, and surrounding the semicircles 

 of bleached cloth on each side of the bar. When the blue cloth with the piece of 

 iron on it rested on several folds of wet blotting-paper, the insoluble blue colouring 

 matter was disti-ibuted on and through the blotting paper in lines which were 

 more distinct and delicate than those oa the cloth itself. These lines of colour were 

 evidently formed under the influence of some force of the nature of magnetism. 



Small horse-shoe magnets were laid on the wet blue cloth, resting on wet 

 blotting-paper, and well-curved lines of coloiu-, usually more or less circular, were 

 developed from the poles and from other parts of the magnets ; some parts of the 

 cloth, usually in the vicinity of the poles, being bleached, and the colour so removed 

 deposited in lines around the bleached portion. 



This action, I found, was not a purely magnetic one, because on placing a piece 

 of gutta-percha tissue over the wet blue cloth, and a magnet on the top of the 

 gutta-percha, so as to prevent it from coming in actual contact with the wet cloth, 

 no effect of any kind was produced after many weeks. 



I next tried the efiect of laying a horse-shoe magnet on a piece of wet cotton 

 cloth dyed with aniline, and found that most of the colour was swept away in a 

 semicircle in front of the poles, and it appeared as if the colour had been swept 

 away from the south pole and deposited around the arc of the circle runnino- from 

 the north pole. Thus it is clear that a force of the nature of magnetism is capable 

 of acting on aniline colour under certain conditions. 



To find the action of a current of electricity on the prussian-blue colour, I placed 

 some of the wet cloth on wet blotting-paper, and laid upon the cloth two platinum 

 electrodes connected with a Leclanche battery ; the current was allowed to pass 

 during several days, and the result was that the colour imder both electrodes was 

 slightly bleached, but more so under the one than the other ; whilst one, and only 

 one, line of colour along the outside of one of the electrodes, was transfen-ed to 

 and penetrated six sheets of blotting-paper underneath. 



Lastly, a strip of copper was laid on the prussian-blue cloth, and a weak current 

 of electricity passed through it, to find what action the forces aldn to magnetic, 

 which surroimd the current, would have on the colour. The copper seemed to 

 decompose the prussian-blue underneath it, producing prussiate of copper, which 

 was transferred to the blotting-paper ; but the most remarkable result was that 

 some of the imdecomposed blue colour was transferred to the blotting-paper in a 

 broad, faint band (along one side only), about half-an-inch from the copper band. 



6. On file Specific Refraction and Dispersion of Light hy Liqtdds. 

 By J. H. Gladstone, Ph.D., F.B.S. 



The general conclusions arrived at from observations made on a large number of 

 liquids were as follows : — 



1. The original statement of the Eev. T. P. Dale and the author, that the length 

 of the spectrum (the difference between the refraction of the lines A and H) 

 decreases with elevation of temperature, is fully confirmed. 



2. The original statement that this length of spectrum divided by the density 



/ Mir f^A \ jg jigai-iy^ ijut not exactly, a constant at different temperatures, seems 



to be confirmed by more careful experiments. 



3. It seems to be the general rule that this specific dispersion slightly diminishes 

 \vith increase of temperature. 



4. This specific dispersion is very notably affected by the constitution of the 

 compound body. Thus among hydrocarbons, the change of the refraction-equivalent 

 of the carbon from 5-0 to 6-1 or 8-8, reveals itself by a change in the specific disper- 

 sion far greater proportionally than that in the specific refraction. 



5. AVhile the specific refraction of a compound is the mean of the specific 

 refractions of its constituents, no such simple law holds good with respect to specific 

 dispersion. 



