1916] THE GROWTH OF ETCH-FIGURES 263 
Any such deep corner will be the starting point of a beak, in the form 
at first of a somewhat tapering excavation. A particle of the solvent in 
a position within this, will be attracted by forces in the directions a and 
b, and will tend to follow their resultant in the direction c, and thus 
c 
increase the length of the excavation. The transition from this to 
a normal beak can readily be conceived. 
(d) The tubular beaks. 
The processes which are here described as tubular beaks, and which 
consist of capillary tunnels running in various directions throughout 
the crystal, might be supposed to be the result of the enormous develop- 
ment of notmal beaks. But this cannot have been their mode of origin, 
for the action described in the previous section could not take place in 
such a narrow space. Many of them indeed are apparently not con- 
nected with pits. Their growth is probably due to the ease with 
which corrosion can take place parallel with a cleavage plane, and 
which is so pronounced that such directions have been called ‘‘solu- 
tion planes”’. 
Several examples were observed in which two pits were joined by a 
connecting tube which evidently owed its existence to such a cause. 
If conditions were such that the tubes could develop along these planes 
the result would be a number of cavities arranged in a definite crystallo- 
graphic direction. But in the most pronounced cases the tubes were 
not confined to any plane. Their irregularity must have been due to 
their deflection from their ordinary course by coming in contact with 
successive points of weak cohesion. 
This phenomenon is of interest in view of modern theories of the 
origin of schillerization. For many years this has been known to be 
