ON DIMORPHOUS BODIES. 183 
ture will appear on comparing those of almost any pair or 
group of isomorphous bodies. ‘The molecules, therefore, of the 
analogous compounds, even of bodies which may replace each 
other, are often separated by unlike spaces. 
Now in two isomorphous substances exhibiting this differ- 
ence, the increased distance of the molecules in the less dense 
may either be equal in every direction, in which case, though 
the densities are not related as the equivalents, the crystalline 
form and dimensions of each would remain alike, or the in- 
crease of distance may be different in the direction of the 
several axes of the crystal, in which case the angular dimen- 
sions of the two substances in a state of crystallization would 
more or less vary. 
Heat is known to expand regularly crystallized bodies un-. 
equally in different directions, enlarging the acute angles and 
imparting a tendency towards the cube or other forms belonging 
to the regular system. The suggestion of Mitscherlich is, that 
chemical affinity acts in the same way as heat, drawing in or 
binding together the molecules more closely in one direction 
than another, so that if, at the temperature at which two isomor- 
phous compounds crystallize, the affinity between the elements 
in the one be only a small degree greater than in the other, a 
difference more or less great must result between the dimensions 
in the so-called plesiomorphous bodies, that is, the crystals must 
be plesiomorphous only. And this suggestion is the more probable 
inasmuch as it accounts for the fact that the plesiomorphous dif- 
ferences do not prevail equally among all the analogous com- 
pounds of the same acids or bases; the difference between the 
affinities of two bases, A and B, for an acid C, being probably 
unlike, not only in amount, but in sign*, to their difference for a 
second or third acid D or E. 
The close relation which exists between chemical affinity and 
heat would predispose us to receive with favour the hypothesis 
in question ; but we can so far test it by observation, since it 
implies that in any isomorphous group those substances whose 
crystalline dimensions most closely approximate should have 
their densities also most nearly in the ratio of their atomic 
weights ; and conversely, those which have the acute angles of 
their crystals the greatest, should also have their densities fur- 
thest below what this ratio would indicate. 
* In the difference (of the affinities?) of baryta and strontia for the sulphu- 
ric and carbonic acids, we have this disagreement both in quantity and in sign. 
In the Rt. Rh. Prisms of these substances we have the obtuse angle in 
Sulphate of baryta = 101°-42! Carbonate of baryta = 118°:30! 
———_ strontia = 104™00! ——-_ strontia = 117°-32’. 
