Bartow—A Mechanical Cause of Homogeneity of Crystals. 565 
ing process. For in order that the arrangement may take place as 
a consequence of the interaction of the parts which are taking up 
symmetrical situations, there must, it is manifest, be a more or less 
extensive tract of matter in an unlinked or partially unlinked, i.e., in a 
Sluid condition, in which symmetrical arrangement is being achieved, but 
not rigidity. There must, in other words, be a state, which may, 
however, be quite transitional, which is a symmetrically arranged fluid 
state. 
Again. Itfsuch a fluid tract be bounded on one side by a portion 
of the same assemblage, of the same composition, which has passed 
into the completely rigid state, but on the other sides is not so 
bounded : it is evident that at the side next to the rigid portion, 
if the latter have retained the same or nearly the same density, 
the parts or particles will, as they pack closely, take up positions 
in harmony with those of the particles composing the rigid portion, 
and will consequently here arrive sooner at a condition of stable 
equilibrium, and so experience less continual disturbance or re- 
arrangement than those do which are further removed from the 
solid portion. 
On the other hand, if the fluid tract be in contact with a 
solidified assemblage which is wnhomogeneous, it is evident that the 
condition of special tranquillity which prevails where the fluid and 
solid meet when they are congruent will not be experienced. 
As in harmony with these conclusions, one may cite the 
existence of crystalline liquids (krystallinische Fliissigkeiten), 
d.e., of a liquid state of some bodies in which a definitely arranged 
symmetrical structure, as evidenced by their polarizing properties, 
is associated with complete fluidity.!. And the rarity of this phe- 
nomenon can hardly be said to diminish its suggestiveness. For 
a single well-established instance of the production of crystal 
properties apart from rigidity suffices to show that the latter is 
not essential, and may in all cases follow, and not accompany, the 
arrangement of ultimate parts which gives crystal properties. 
Parallel to the above conclusion that if, in a homogeneous 
1 Lehmann, ‘‘ Ueber fliessende Krystalle.”’ Zeitschr. fiir physikalische Chemie 4, 
p. 462 ; and by thesame author, ‘‘ Die Struktur krystallinischer Flissigkeiten.’’ The 
same Zeitschr., 5, 427. 
