bd 
7 
566 Scientific Proceedings, Royal Dublin Society. 
assemblage of mutually-repellent particles, a portion have passed 
into the rigid state, we shall have a tract or film of the portion 
still fluid lying next to the solidified portion, and which is in @ 
more tranquil state, and further advanced towards stable equili- 
brium than the rest, we have the fact that it is the portion of the 
crystallizing liquid nearest to the crystal which commonly passes. 
to the solid state first. In other words, crystals grow by accretion, 
while amorphous bodies do not. 
The tract of maximum tranquillity and symmetry which we 
have concluded must lie next to solidified portions of the assem- 
blage will, it is evident, to some extent be broken up if the con- 
tinuity of structure of the solidified surface is partially destroyed 
by rupturing the solid and disarranging the broken portions. For 
a symmetrical arrangement of the film of liquid in harmony with 
one portion of the body will not then be in harmony with a 
neighbouring portion. A very significant fact with regard to 
crystals may be cited to compare with this. 
Tf a crystal, e.g., of ammonium chloride, be beaten out, and then 
placed in its solution and allowed to grow, it is found that the 
accretion at the boundary, where the solid particles are in contact. 
with the solution, takes place in skeleton form? showing that the 
disturbing influence of a fracture and displacement on the process of 
crystallization eatends to a considerable distance from the fracture im 
both directions in the adjacent film of solution. | 
The transition state, in which the symmetrical arrangement, 
but not the solidification of an assemblage is taking place, is. 
1 The alternative theory of crystal formation supposes each particle added to the 
growing mass to take up its symmetrical situation not before, but at the time it is. . 
attached to the previous growth, just like a brick added to a wall in building. This. 
view seems perhaps at first sight to be countenanced by the discovery made by Wulff 
that better and more regularly formed crystals can often be obtained when a solution 
is in motion that when it is at rest. That it is not supported by the fact referred to is. — 
seen, however, when we consider that the places of minimum disturbance will still, 
notwithstanding the motion, be close to the crystal surfaces, and that while the 
disturbance operates to prevent irregular growth, it does not preclude the existence of 
Jilins of the crystallizing substance in a more or less liquid condition adhering to the 
already solidified crystal surfaces from time to time formed, and remaining undisturbed by 
the motions. (See Wulff, Zeitschr. fiir Kryst., 11, p. 120. 
* Lehmann ‘‘ Ueber fliessende Krystalle’’ Zeitschr. fiir physikalische Chemie 4, 
p. 467 and fig. 3. 
