1903.] Heat and of Solvents on Thin Films of Metal 233 



If the foregoing views are correct, it will follow that a freely 

 suspended aggregate of solid molecules will take the spherical form 

 if its surface is sufficiently large compared with its mass. A globule 

 will only take the external form of a crystal when its mass becomes 

 large enough to permit the crystallic force to overpower surface 

 tension. The researches of Neville and Heycock* have shown that, 

 even in the solid state, the larger crystals grow at the expense of the 

 smaller and eventually swallow them up. The range of the crystallic 

 force therefore widens as the mass of the crystal is increased, while 

 the range of the molecular forces remains constant. It seems 

 possible, therefore, that the average size of the granules in solids 

 may result from the establishment of a state of equilibrium between 

 these rival forces. 



The bearing of these views on the theory of the formation of 

 globular precipitates and the sphero-crystals described by Quinckef is 

 obvious. The oily drops of calcium carbonate, which are seen as 

 a first stage in the development of crystals, are not necessarily 

 a super-saturated solution of calcium carbonate, they may be aggre- 

 gates of solid molecules under the control of surface tension. 



Mr. F. H. Neville has directed my attention to Faraday's observa- 

 tions I on the fluidity of minute globules of sulphur at the ordinary 

 temperature. These observations seem to support the views I have 

 advanced here. 



Metals slowly deposited from solution tend to aggregate in the 

 granular form. If the deposition takes place on nuclei floating in 

 the solution or resting on the submerged surfaces, spheroidal granules 

 are produced. If the deposition takes place on a clean smooth surface 

 a thin flat film is produced, but even then the deposit is made up of 

 flattened granules or " spicules " as I have called them in a former 

 paper. When the thin film is annealed by heating, the flattened 

 granules are drawn up into more distinctly rounded forms. 



The flattened granule or "spicule" can be detected by oblique 

 illumination even when it is too thin to show by transmitted light, 

 By comparing its size and appearance under different forms of 

 illumination when it is sufficiently massive to show by transmitted 

 light, we are enabled to feel some confidence in the reality of those 

 appearances which can only be seen by oblique illumination. 



In the separation of solids directly from the gaseous state, e.g., 

 lamp black or magnesium oxide, the granular form is assumed by tho 

 molecular aggregates. 



The more detailed study of these granular and spicular appearances 

 will be dealt with in a further communication. 



* ' Roy. Soc. Proc.,' vol. 69, p. 325. 



t ' Ann. d. Physik,' vol. 7, pp. 631682, 1902. 



' Quarterly J. of Science,' vol. 21, p. 392. 



