J. D. Dana’s Mineralogical Contributions. 87 
present. It is obvious that crystallization must follow, or go 
hand in hand with composition, but not lead the way in a classi- 
fication of inorganic substances. 
Hermann has endeavored to meet some of the difficulties pro- 
duced by homeomorphism, by supposing two or more primary com- 
pounds as the bases of a class of species, the occurrence of whic 
in different proportions shall form all the species of an homeo- 
morphous group. This has an appearance of propriety as regards 
some groups of Silicates. But what symmetry in the constitution 
of the two can be made out in this way when Heavy Spar and 
Graphic Tellurium are compared; or Sulphur and Scorodite ?— 
or Arragonite, Bournonite and Nitre ?—or Chrysolite and Epsom 
salt ?—or Pyroxene, Glauber salt and Hydromagnesite ?—or Brook- 
ite and Columbite ?—Such facts evince that homeomorphism is 
seem to matter not what the elements are ; if only the resultant 
has a certain relation as regards atomic volume to the atomic vol- » 
ume of another compound, there is isomorphism. 
Still there is often in related groups, a numerical relation in 
the elementary constitution which affords an explanation of the 
atomic volume relation, without looking to other considerations : 
and this numerical relation may be extended to the whole, when 
n this way, Avinite and Danburite have a like relation and both 
are triclinic. In axinite the ratio between the oxygen of the 
bases and boracic acid, and that of the silica is 1:1; and the same 
Is true for Danburite as the recent analyses of Smith and Brush 
Show.* The formulas may be written, for axinite (R*,#, B,) Si; 
for Danburite (Ga?, B,) Si. The propriety of reckoning the bo- 
racic acid with the bases is shown by the fact that in Tourma- _ 
line the ratio thus obtained is the only one that is constant for all 
* This Journal, [2], xvi, 865. 
