MINERALOGY, 65 
Alkali (by difference), . .  . 417 1.48 
Ignition, . . e . . “71 1.17 
100.00 100.00 
Here, again, the same results are evident, the preponderance of alumina 
in the hornblende being the most striking difference. 
But in the case of the pyroxene at Edenville, and also in some of our 
igneous and metamorphic rocks, the pyroxene has changed its cleavage 
and optical properties, and become hornblende. The minerals have not 
become hydrated, or the iron oxidized, as in ordinary decomposition ; but 
a change has been effected without any alteration of composition, form- 
ing the mineral which was called by Rose uralite, which is pyroxene 
having the inner structure and optical properties of hornblende. In 
this case it is evident that pyroxene and hornblende are dimorphous 
forms of the same composition; that is, the molecules of the original 
pyroxene, under the subsequent influences, have rearranged themselves. - 
Tschermak has noticed that crystals of pyroxene with crumpled ends 
are the ones most liable to this kind of alteration. This is the case with 
the Edenville crystals, but whether it be a result or a cause of the altef- 
ation is not plain. 
From these analyses, it appears that any given composition capable of 
forming pyroxene, may, under different circumstances, form hornblende, 
and hence the two minerals are dimorphous forms of the same material ; 
but that under uniform conditions, chemical composition will atten 
what species shall be formed, and alumina is an important agent in this 
determination. 
Now, in the decomposition of rocks and the redeposition of sediments, 
the lime that is held in their composition is most apt to be dissolved and 
carried away in solution, and thus we-obtain those immense beds of lime- 
stones, whereby the sediments are left more aluminous ; and so in meta- 
morphic rocks there is a much greater tendency to the formation of 
hornblende; and our stratified basic rocks are mostly diorites and amphib- 
olites, while the amount of metamorphic pyroxenic rocks is small. As 
confirmation of this, see the following analysis of hornblende from the 
diorite of Littleton: 
VOL. Iv, 9 
