1903.] Heat and of Solvents on Thin Films of Metal. 227 



As a further result of this study, it is now suggested that the 

 behaviour of surface films during heating may be most satisfactorily 

 explained on the hypothesis that even at a temperature much below 

 the melting point, sufficient freedom is conferred on the molecules by 

 the heating to enable them to behave as the molecules of the liquid 

 metal would do, and to arrange themselves under the influence of 

 surface tension either in films or in drop-like or granular forms. 



Faraday definitely associated diminished transparence and enhanced 

 reflecting power with a state of strain induced by pressure and 

 burnishing. My later observations confirm and extend this conclusion 

 and show that, contrary to my first impression, all mechanical dis- 

 turbance of the surface by polishing or burnishing tends to diminish 

 transparence and to increase the reflecting power. 



In gold an olive green colour by transmitted light is generally 

 associated with the strained condition, while a great variety of shades 

 of colour, ranging from ruby-red to violet, are found in the annealed 

 metal. 



A gold leaf, after annealing on glass, retains all the ribs and 

 markings of the original leaf. These interfere with the uniform 

 adhesion between the metal and the glass, and lead to obvious breaks 

 in continuity. Fig. 13 (Plate 11) shows a piece of gold leaf annealed 

 on glass. Small granular aggregations are seen, but the continuity 

 of the film between these is not so obvious. 



Faraday describes how he prepared gold films of varying thickness 

 and of great smoothness and continuity by placing specks of phosphorus 

 on the surface of a dilute solution of gold chloride. 



Films made in this way were annealed by heating on glass. 

 Figs. 15 to 18 show the effects of annealing on films of various 

 degrees of thickness. 



Before annealing, the thicker films (figs. 15, 16, 17) were olive- 

 green by transmitted light and had a full yellow colour and metallic 

 reflection by reflected light. The thickest film, fig. 15, required a, 

 very intense light to show its green translucence. Fig. 16 had the 

 same depth of colour as gold leaf, and fig. 17 was thinner and paler. 

 Fig. 18 was much thinner than any of the others, showed hardly 

 any metallic reflection and its colour was blue-purple by transmitted 

 light. 



After annealing the colour of this film was rose-pink by transmitted 

 light. The photograph shows that it was quite continuous, but with, 

 thickenings or aggregations distributed regularly over the surface.- 

 The photograph has been made somewhat dark in order that the 

 structure may be shown more distinctly, but as seen under the micro^ 

 scope the surface was entirely free from dark or opaque patches. 



In fig. 17 the retraction into globular or rounded forms of greater 

 opacity is very pronounced, but the continuous transparent film 



VOL. LXXII. R 



