Bartow—A Mechanical Cause of Homogeneity of Crystals. 669 
tion, their faces becoming crooked or approximating to the 
spherical. 
To compare with the conclusion reached above that the re-in- 
troduction of abstracted portions of an intercalated assemblage 
may restore internal regularity and anisotropism, if the latter 
subsisted ; we have the fact that in some cases imbibition appears to 
increase the regularity of the internal arrangement of the parts of 
a crystalloid. Thus when crystalloids of musa or sparganium 
swell up in pure water the double-refraction materially increases. 
In the case of the former the phenomenon is very striking, for 
placed between crossed nicols when dry, or with oil, the light trans- 
mitted is very faint, but when so placed after imbibition it shows 
a brilliant white; and if a quartz plate is used which gives red, 
we get instead of the slight change of tint produced by the crystal- 
loid before imbibition, vivid yellow, white, blue-green, or green 
according to the position.’ 
The permeability of the sponge-like framework by a given 
set of constituents will, according to the above, depend principally 
on whether their entry will increase the closeness of packing, and 
in this connection may be mentioned the experimental observation 
of Tammann that what determines whether or not the respective 
molecules of a number of liquids shall penetrate a given membrane 
is not their relative size, as has been supposed by Traube and by 
Ostwald. 
We have said that when the framework is squeezed together 
the maximum distortion of parts will take place near the outer 
boundaries, because the conditions are there least symmetrical. 
From the same cause some parts of the shrunken framework will 
be more prepared to pack closely with the incoming liquid than 
will other parts, the distribution of this difference being dependent 
on the outer form of the mass of the framework. 
To compare with this we have the fact that if erystalloids 
of musa hilléi are treated with weak acid, so that a minimum 
1 [bid., p. 153. 2 Ibid., p. 154- 
3 Tammann, ‘‘ Ueber die Permeabilitat von Niederschlagsmembranen.’? Nach- 
richt der Géttinger Gesellschaft der Wiss., 1891, p. 213. Comp. Walden, “ Ueber 
Diffusionserscheinungen yon Niederschlagsmembranen.”’ Zeitschr, f. physical. 
Chemie, 1892. Bd. x., p. 699. 
