Prof. A. R. Richardson, D.S.O., F.R.S.

regions on the surface, where there are many kink sites close together, so that silver atoms released by the process described above can easily combine. These rough regions are the sensitive parts of the crystal ; the work of Mitchell's group shows them to occur, as one would expect, where a substructure or array of dislocations meets the surface. This, at any rate, is what happens in large crystals, and probably in grains of emulsions, too. A model of this kind seems to fit many of the facts and probably comes nearer to the true mechanism for many actual emulsions than the original Gurney-Mott theory. It is remarkable how an understanding of the photographic process seems to involve all the latest concepts of the theory of solids, from the movement of excitons to diffusion of atoms along dislocations. N. F.