2. Iledenbergite consists chiefly of silica, lime, and protoxide of iron, as may be seen by the following analysis by G. Rose of a variety from Lunaberg : Silica Lime . 20'87 Protoxide of Iron . . . . . . 26'08 Protoxide of Manganese with Magnesia . . . 2'98 98'94 It is of a dark-green colour, sometimes nearly black.
3. Saldite, those varieties in which the magnesia is only in part replaced by protoxide of iron, and which may be regarded as consist ing of one equivalent of Iledenbervite united with two of Diopside. A variety is called Malakolith. (` Anwcndung der Lothrohrs,' by Berzelius.) 4. Diallage. The difference in the analysis by Kohler of two speci mens, the first from Tuscany, the second from Ulthenthal in the Tyrol, were as follows : Silica Lime . . . . . 19'08 2.19 Magnesia 14'91 29'67 Protoxide of Iron . . . 8'67 8'46 Protoxide of Manganese . 0.38 Alumina . . . . . 2'47 2407 Water 117 0'21 — — 100'02 This variety is characterised by its mother-of-pearl lustre, and by its possessing the most perfect cleavage iu the direction of the diagonal of the prism. It is seldom found in perfect crystals. Its most general colour is a bronze yellow.
5. Ilypersthene, which is very similar in its appearances and characters to Manage, is a bisilicate of iron and magnesia, Both of the last-mentioned varieties may be distinguished from the former, as well as from each other, by means of the blowpipe, and by attendiug to the following characters as stated by Berzelius :— Diallage alone in a matrass decrepitates, becomes of a lighter colour, and gives off a little water.
On charcoal it is with difficulty melted on the edges into a gray scoria.
With borax it is with difficulty fused into a clear glass, somewhat coloured by the protoxide of irou.
It is decomposed by the phosphate of soda and ammonia, with the development of the silica.
Ilypersthene, on the contrary, when heated alone iu the matrass, decrepitates slightly, gives out a little water, but does not change its appearance; while on charcoal it readily forms a green opaque glass, as is also the case when heated with borax.
The salt of phosphorus does not apparently decompose it, but the mineral at first becomes rounded on the edges, and may at length be entirely fused.
The structure also deserves particular attention, the cleavage-planes in ypersthene being perfect, both in the direction of the faces r and M, the latter of which are obtained in Dialloge with very great difficulty.
We have now described the various species generally considered as comprehended within the genus A agile or Pyrozene ; but Professor Gustave Rose has endeavoured to prove the necessity of uniting Augite and hornblende (Pyrarene and Amphibole) into one genus. ilia argu ments for this union are the following :-1Ie first shows that the two prisms of A vim and hornblende, however different in appearance, admit of being derived the one from the other, according to the laws observed to connect the crystallographic forms of varieties of the same genus in other minerals. To show this, let the accom panying parallelo gram, whose semi diagonals are a and b, represent the horizontal section of the prism of Augite; since the whole angle of this prism at A is 87° 6', b is the tangent of an angle of 43° 33'; if this tangent be doubled, the corresponding angle will be found to be 62' 15' 25', the double giving 124° 30' 50', an angle agreeing most closely with 124' 31', the angle obtained by Mitseherlich in a species of hornblende when measured by Wollaston's reflecting goniometer. The larger parallelogram, therefore, formed by doubling the diagonal b, is the horizontal section of the prism of hornblende.
A similar relation is also approximately true for the inclination of the faces s in Augite and r in Hornblende; for if the angle 120° 57' of ./1 ugite be halved, and its tangent doubled, the corresponding angle is 11' 21', and by doubling this we obtain 148° 22' 42", not much differing from 14S° 25', as found between r in hornblende of Vesuvius by Rose.
His argument drawn from the chemical constitution of these minerals is by no means so satisfactory; for though in hornblende we find a series of bisilicates of the same .bases, and ns it were running parallel with those • already described as A ugites, the circumstance observed by Bonsdorff, that all the varieties of hornblende contain fluorine, while Gustave Rose has been unable to detect that element in A ugite, weakens the connection between these minerals, and renders the determination of what part the fluorine acts in their constitution a most desirable object. Our ignorance on this point, however, and the difficulty of determining what is the action of the alumina, which occurs in considerable quantity in some ornblendes, prevent us from forming any opinion from the results of chemical analysis.