Bartow—A Mechanical Cause of Homogeneity of Crystals. 673 
amorphous, 7. e., destitute of any qualities depending on direction, 
but at the same time it will present general uniformity of distribution 
of the constituents. We shall indeed picture a mass of similar 
curved assemblages, just like a single continuous assemblage, as 
continually striving after that relative arrangement of the parts 
composing it in which it occupies least space, but that owing to 
inability to reach stable equilibrium it continually fluctuates from 
one imperfectly symmetrical closest-packed arrangement to another 
as itis shaken up by passing disturbances. The different consti- 
tuents will be very evenly distributed and a like relation of parts 
will frequently be found repeating itself in various places in the mass. 
If a collection of very minute assemblages thus circumstanced 
be solidified we shall look for the production of a practically homo- 
geneous amorphous solid, destitute of any appreciable symmetry, 
and therefore paralleled by the amorphous or non-crystalline state 
of bodies. 
It should be remembered, however, that we have concluded that 
some thin curved assemblages are convertible by accretion, under 
favourable conditions, into homogeneous assemblages,’ and it is 
conceivable that this may always or nearly always be the case. 
With this we may compare Lehmann’s conclusion that the amorph- 
ous state is due to hindrances which prevent the molecular forces 
from achieving the degree of symmetry of arrangement which they 
would otherwise attain.’ 
The next effect to be considered is :— 
V.—Combination of two or more homogeneous or approximately 
homogeneous assemblages to form a single homogeneous or 
approximately homogeneous assemblage. Resemblance of 
this effect («) when in its most perfect form, to that highly- 
symmetrical intermixture of the combining atoms or com- 
plexes which must, it is evident, accompany or precede a 
chemical synthesis; (4) when in its. less perfect form, 7.c., 
where the assemblage produced is only imperfectly homo- 
geneous or is fluctuating, to some phenomena of diffusion. 
If in an assemblage composed of one or more different kinds of 
1 See p. 572. 
2 See Lehmann in Zeitschr. f. eee 1, pp. 116, 117, and comp. Pope’s translation 
of Fock’s ‘‘ Chemische Krystallographie,”’ p. 61. 
302 
