TRANSACTIONS OF SECTION A. 653 



strong, the effect of the descending current was to promote the thinning of the 

 film. An ascending current, on the other hand, checked the thinning of the film, 

 retarded the formation of coloured rings if applied from the beginning, and rapidly 

 caused the disappearance of the black if applied after a ring of black had already 

 been formed. These effects were more marked with the soap solution than with 

 the ' liquide glycenque,' but the nature of the effect was the same with the two 

 liquids. The results are valuable, as giving the means of controlling the thinning 

 of a film in cases where it is desired to compare the properties of two films of 

 different thickness. 



7. On the Diffusion of Metals. 

 By Professor W. Chandler Roberts, F.E.S. 



8. Gil some Phenomena connected tuith Iron and other Metals in the solid 

 and molten states, with notes of experiments. By W. J. Millar. 



(1.) Object of Paper. — Results of experiments by the author with various metals, 

 such as cast-iron, gun-metal, phosphor bronze, lead, copper, and type-metal. The 

 object being to determine the cause of the well-known phenomenon of the float- 

 ation of cold cast-iron on molten cast-iron, and as to whether any expansion took 

 place upon solidification in the metals above noted. 



(2.) Notes of some of the experiments from which the author concludes that 

 the cause of floatation of the solid metal on liquid metal of the same kind is 

 buoyancy, due to expansion suddenly set up in the immersed pieces, and that this 

 expansion was found by careful measurement to be at least equal to the shrinkage 

 or total decrease in length of the piece from white hot solid to finally cooled down 

 solid. 



Further, that the expansion observed is obtained within much lower limits of 

 temperature than the shrinkage; as the pieces, which were in all cases removed 

 from the molten metal, immediately on appearing floating hardly showed redness, 

 and when broken, it Avas found that, the crystalline character of the metal 

 remained. 



(-3. ) Notes of experiments made by gradually heating pieces of cast-iron — the 

 results of all these experiments leading the author to conclude that the rate of 

 expansion in cast-iron is at first much more rapid at low temperature than after- 

 wards at high temperature. 



(4.) From experiments carried on with pieces of lead, copper, and type-metal 

 it was found that if any floatation occurred, it was only with small light pieces — 

 heavy pieces sinking and remaining at bottom of ladle. 



Gun-metal and phosphor bronze behaved like cast-iron. 



(5.) Consideration of the peculiar appearance, or ' break] observed on the 

 surface of molten cast-iron, the figures presenting a geometrical pattern, like inter- 

 lacing circles or stars. 



The author believes that this appearance is due to cracks forming upon the 

 rapidly forming skin — these cracks taking more or less a circular form from the 

 convex forms, into which the various parts of the surface are thrown, due to the 

 bubbling up of gas or air. 



This appearance is limited to cast-iron, and experienced observers can tell the 

 quality of the iron from the form of pattern or figures showing on the molten surface. 



(6.) From observation and experiments carried out from time to time, the 

 author concludes that no perceptible increase of volume of the metals noted occurs 

 ftt the moment of solidification, at least when free from air or gas confined within 

 the casting. 1 



c 



9. On the Velocity of Light of different Colours. 

 By Professor George Forbes. 



-ee nlso Proceedings, of the Royal Society of Edinburgh, Session 1881-82. 



