682 Scientific Proceedings, Royal Dublin Society. 
water considerably lowers the rate of diffusion of a diffusing 
body.? | 
The action of the principle of closest-packing may cause an 
assemblage which consists of, or contains a linked framework’ to 
exhibit a selective action on assemblages brought in contact with it. 
Thus for closest-packing to obtain, a certain set of balls may have to 
occupy the interstices among the balls of such a framework, and 
to quit the interstices of another assemblage in contact with but 
outside it. 
In this connection we may recall the phenomenon of the re- 
moval of dye-stuffs by the action of crystalloids. Fuchsine blue, 
in particular, is greedily taken up by many crystalloids, so that 
in twelve hours a brightly-coloured solution of this substance may 
be rendered entirely colourless by the action of these bodies lying 
ro Tt 
We come now to :— 
VI.—The breaking up of an assemblage into two or more distinct 
assemblages ; an effect which resembles the disentangling of 
the separating atoms or complexes that commonly follows a 
“‘ chemical decomposition,’ and also resembles the crystal- 
lizing out of a constituent from a liquid or partially liquid 
mixture. 
If an assemblage composed of two or more different kinds of 
units, whose arrangement at the time of its formation was the 
closest-packed possible, is subjected to a change of conditions such that 
this arrangement ceases to be so, it is evident that, if the state of the 
assemblage as to linking permits, it will in obedience to the fun- 
damental law of closest-packing break up into two or more assem- 
blages of different kinds. 
The re-arrangement of the parts to form these separate assem- 
blages will be such as to give closest-packing under the new con- 
ditions, and these assemblages will be homogeneous or amorphous 
at the behest of the principle of closest-packing. 
———s 
1 «Untersuchungen tiber Diffusion von in Wasser gelésten Stoffen.’’ Zeitschr. f- 
physikal. Chemie, 1892, Bd. x., p. 51. 2 See p. 666. 
° Schimper, ‘‘ Ueber die Krystallisation der eiweissartigen Substanzen.’’ Zeitschr. 
f. Kryst., 5, p. 156. 
