VOL. LViri.] PHILOSOPHICAL TRANSACTIONS. 513 



of which however may be wiped off with a feather, so as to show the colours 

 under it. An attempt to wipe off more, on the rough side of the steel, took off 

 the colours along with it ; but more than half yet remains, with the dust upon 

 it, as it was first formed. 



9. It makes no difference whether the electric matter issue from the pointed 

 body upon the plate, or from the plate upon the pointed body ; the plate op- 

 posed to the point being marked exactly alike in both cases. Also the points 

 themselves, from which the fire issues, or at which it enters, are coloured to a 

 considerable distance, often about half an inch, but not very distinctly. The 

 colours likewise return here, in concentric rings, as upon the plate. 

 • 10. The more circles that are made at the same time, the more delicate will 

 the colours be ; whereas the surface is, as it were torn, or corroded by more 

 violent explosions ; which makes the colours appear rough and course. But this 

 roughness is only perceived on the steel. On silver, tin, and polished brass, the 

 colours were always free from that roughness. 



1 1 . A polished surface is not necessary, the colours being very manifest on the 

 rough side of the steel, where it is not covered with the black dust mentioned 

 above. 



12. These coloured rings appear almost equally well on all the metals on which 

 he made them ; viz. gold, silver, copper, brass, iron, lead, and tin. 



13. When the pointed wire was made to incline to the plane on which the 

 colours were exhibited, the circular spot was quite round, the centre of it being 

 in the perpendicular let fall from the point ; but the colours were projected 

 opposite to the point, in an oblong figure. 



On showing these coloured rings to Mr. Canton, Dr. P. was agreeably sur- 

 prised to find, that he had likewise produced all the prismatic colours from all 

 the metals, but by a different electrical process. His method had been to extend 

 fine wires over the surface of pieces of glass ; and when the wire was exploded, 

 he observed that the glass remained tinged with all the colours from all the 

 metals. They are not indeed disposed in so regular and beautiful a manner as in 

 the rings above produced ; but they equally demonstrate, that none of the metals 

 discovers the least preference to any one colour more than another. A variety 

 of other very extraordinary apj)earances occurred in the course of Mr. Canton's 

 experiments in melting wires. 



In what manner these colours are formed, it may not be easy to conjecture. 

 In Mr. Canton's method of producing them, the metal seems to be dispersed in 

 all directions from the place of explosion, in the form of spheres, of a very great 

 variety of sizes, tinged with all the variety of colours, son)e of them too small to 

 be distinctly visible by any magnifier. In Dr. P.'s methotl, it should rather seem 

 that they are produced in a manner similar to the production of colours on steel 



VOL. XII 3 U 



