Sullivan — On Augitc and Ilornhlende. 
41 
hand, being formed by the successive additions of molecules of the 
primary meta-silicate — that is, any condensed meta-silicate is a simple 
multiple of the first. The relations of the three classes of silicates may 
accordingly be represented as follows, taking as the example the sili- 
cates containing six atoms of silicum in each class : — 
Ortho-silicates. {M"7Si60i9 = (M^SiOs)^ + M^^SiOi 
Meta-silicates. {M'eSieOis = (M"Si03)^ 
f I. M'gSieOn = (M"Si03)5 + SiOg 
ii\ II. M"4Si60i6 = (M"Si03)* + 2Si03 
'^1 <; III. M"3Si60i5 = (M"Si03)' + SSiOz 
This table expresses very clearly what I have already stated, namely, 
that the condensed meta-silicates, being formed by the association of 
similar simple molecules, the condensed molecules must be isomorphic 
with the constituent simple molecules ; while both the condensed ortho- 
silicates and condensed anhydro-silicates containing the elements of two 
different kinds of molecules, change with each degree of condensation. 
The preceding table also explains the presence of free quartz in 
rocks composed of anhydro-silicates, such as potash or soda feldspars, 
which belong to this class of silicates ; and the almost complete ab- 
sence of quartz from rocks consisting chiefly of meta-silicates, or where 
they are present in any quantity, as in syenites, dolerites, &c. It also 
throws much light upon the mutual relation of the different minerals 
which are found asssociated together in nature, — a subject which I 
cannot now discuss. 
I now come to the subject of the constitution of the aluminous 
hornblendes and augites. Both these classes of minerals contain in 
addition to the meta-silicates of the dyad metals, aluminum and, gene- 
rally, one or both of the alkaline metals, potassium and sodium — the 
latter as silicates — and part, at least, of the aluminum also in that state. 
Crystals of aluminous augites or hornblendes always present uneven 
faces of cleavage, and the faces are frequently drusy, and weather ir- 
regularly — qualities which are characteristic of all heteromorphic 
crystals — that is, of crystals built up of different compounds, crystal- 
lizing separately in different crystalline series, but which, when 
mingled together, follow the law of one of them. These properties led 
Kenngott'^* and Frankenheimf to the conclusion that aluminous augites 
and hornblendes were mixtures of the pure minerals with alumina or 
some aluminous compounds. Is the alumina always present in the same 
state of combination — that is, does aluminous hornblende or aluminous 
augite always contain the same aluminous silicate ? Again, is there 
only one such definite silicate ? And, again, are the aluminous silicates 
of the minerals of the hornblende and augite groups the same or dif- 
ferent? The whole problem of the constitution of the aluminous 
augites and hornblendes, with the exception of one point, is embraced 
* Sitzungsber. d. "Wien. Akademie, xii. 702. f Poggend. Annalen. xev. 375. 
R, I. A. PROC. — VOL. I,, SER. II., SCIENCE. G 
L v. M" SieOia = (M"Si03j + SSiOj 
