and Soft States in Metals. 267 
smooth glassy layer, the surface of which is either as featureless 
as that of a liquid, or is drawn into smooth ridges and furrows 
which show unmistakable signs of having been ploughed in a 
viscous liquid filled with granules. (Figs. 5-14.) 
40. By polishing across the ridges and furrows which have 
been ploughed by fine emery, a smooth covering layer is 
spread alike over ridges and furrows. By arresting this 
cross flowing at various stages its progress may be studied, 
and photographic records may be preserved which confirm 
the view that an essential feature of all polishing is the 
transient existence of a layer of molecules in a state of mobility 
closely analogous to that of a liquid. Records of this kind 
have already been published *. (Figs. 4-12.) 
41. By a comparison of the structure which results from 
polishing with that which is developed by beating, rolling, or 
pressing, it is found that those effects which in polishing are 
strictly confined to the surface, penetrate to a greater and 
greater depth under the more drastic forms of treatment. 
It is thus made plain that the effects of flow are not confined 
to the outer surtace only, but that they take place at all points 
where the strain reaches the stage at which mobility of the 
molecules is induced by the movement of one portion of the 
substance against another. 
42, Accepting the existence of the mobile phase as proved 
by the records of micro-structure, the complete transformation 
from soft to hard may be written C-——> M——->-A;; M being 
the intermediate mobile phase. 
43. The condition which approaches most nearly to that of 
this mobile phase is that existing in an under-cooled liquid in 
which molecular freedom continues after the temperature has 
fallen below the solidifying-point. In this case it may be 
supposed that the molecular activity due to temperature has 
suffered curtailment without actually disappearing into the 
comparative fixity of the solid state. Till the temperature 
has fallen below the transition point the fixity is only compa- 
rative ; for we know that at this point there is still sufficient 
molecular activity to complete the transformation from A to UC. 
In the case of silver the transition-temperature is not higher 
than 250°, that is about 700° below its liquefying-point. But 
below the transition temperature the molecules must remain 
fixed relatively to each other, for otherwise it would be 
difficult to account for the state of strain which exists in the 
hardened phase as shown by its optical and electrical pro- 
perties (pars. 27 to 34 and 70). ‘The mobility of the mole- 
cules in the M phase is therefore not due to heat, but to motion 
* Proc. Roy, Soe. vol. xxii. pp. 218-226, figs. 1-12. 
