(Persia, Pennsylvania, Amiano in Parma, Canada, California, &c.).

Asphalte is found in many localities (e.g. at the Dead Sea; Trinidad, where there is a complete pitch-lake; at Poldice in Cornwall it occurs in granite).

An intermediate substance between naphtha and asphalte is elastic mineral pitch or elaterite (Castleton in Derbyshire).

All these bituminous substances are of vegetable or animal origin, partly products of distillation of organic remains. They frequently occur as admixtures in shales and other rocks, which have received the name of bituminous (Autun in France, Bonn, Markersdorf in Bohemia, &c.)

124. Mellite (Mellilite, Honey Stone).-Dimetric, usually in pyramidal crystals, singly imbedded. Cleavage pyramidal, very imperfect. Fracture usually conchoidal. Somewhat brittle. H.=2-2.5. S.G.=1.5-1.6. Colour honey-yellow to waxyellow, seldom white. Lustre resinous. Semi-transparent to translucent. Cp.= Äl(C403)+18H. Bp. it carbonises with smell of burning; on charcoal burns to a white ash, which acts like pure alumina. It is readily and completely soluble in nitric acid.

Mellite occurs as an accessory ingredient in Browncoal (Artern in Thuringia, Luschitz in Bohemia).

CHAPTER II.

ANALYSIS OF ROCKS.

MICROSCOPIC ANALYSIS.

Ir not unfrequently happens that the various mineral ingredients of a composite rock are so small and intimately blended together as to be entirely undistinguishable even to the practised eye unaided by magnifying power. A simple lens will often render great service in this respect, but the aid of magnifying power may be carried much further with the microscope. For the microscope very thin plates of a rock, so thin as to be somewhat transparent, are cemented on glass, and by the aid of a powerful instrument, textures apparently quite compact are frequently resolved into a web of minute crystals, or we find individual crystals become prominent (porphyritic) in an actually compact matrix. The form of these minute crystals is sometimes to be recognised, and so serves as a guide to the determination of the mineral in doubtful cases. If we further call in the assistance of polarised light, we are enabled to pronounce, with greater certainty, on the amorphous or crystalline character of the compact mass, and on the character of the crystals which by these means are brought to view.

Delicate investigations such as these no doubt require the assistance of complicated apparatus and demand time, so that they are quite out of the question for the geologist on his travels; but as we have said, much may be discovered by a simple lens, which for the practical geological purposes of the general inquirer is in most cases sufficient.

MAGNETIC ANALYSIS.

An admixture of magnetic iron-ore makes many rocks magnetic in their entirety, so as to affect the magnetic

needle, or if the iron-ore be present in small quantities only, it may be discovered by abrasure with a sharpedged magnet, the magnetic particles of the powder so formed clinging to the magnet like a beard. As, however, magnetic iron-ore occurs in many very different rocks, its discovery does not often afford much help to the geologist in determining the character of any given rock.

Fostemann and Delesse have made careful investigations of the magnetism of many different rocks. The former is of opinion that by means of careful magnetic experiments, we ought to be able to ascertain whether a rock be of volcanic or neptunian origin, whether it has been rendered metamorphic by heat, whether it has retained its original position or been subsequently displaced (vide Poggendorff's Annalen, 1859, vol. cvi. p. 106). Delesse had previously discovered that almost all igneous rocks were somewhat magnetic as well as many sedimentary and metamorphic rocks. (Annales des Mines, 1849, vol. xv. p. 1, and Bulletin de la Soc. Géol. de France, 1850, vol. viii. p. 108.)

CHEMICAL ANALYSIS.

The geological interest attaching to the chemical analysis of rocks is chiefly in respect of the nature of their origin.

In the early stages of the science the analysis of composite rocks was conducted by mechanically separating, as far as possible, their several mineral ingredients, and analysing each mineral species individually; and this method is still sometimes adopted where the parts are very distinct and easily to be separated. Compact rocks, such as basalt, were mostly considered as simple mineral substances, and so analysed. When, however, it came to be recognised that many apparently homogeneous rocks were but mechanical compounds of several minerals, chemical analysis was directed to the discovery of these mineral constituents too intimately mixed to be distinguished by the eye.

Gmelin introduced the method of treating a powdered mass of rock with muriatic or other acid, and so sepa

rating it into a part soluble, and another part insoluble in such acid. These two parts he separately analysed, and reduced the results into chemical formula. The object he had in view was chiefly to discover the mineral constituents of the rock. But this mode of analysis is inadequate for the purpose, since few minerals are wholly soluble, or wholly insoluble, in acids, and therefore, instead of the several minerals being separated from each other, a part of each is dissolved and a part of each left, and no definite result as to the original structure can be attained. It is found that even the elementary constituents cannot be successfully so divided; but that some elementary substances are only partly dissolved and partly precipitated by the process. Nevertheless, as a rough approximate, and somewhat empirical mode of suggesting rather than proving the constituents of a rock, it is still sometimes employed, and may in certain cases be of use.

As the chemical character of minerals came to be better known, less reliance was placed on chemical analysis as a means of ascertaining and distinguishing the mineral ingredients of rocks. A small number of elements are so universal in their character that they enter into the composition of a very large proportion of the whole series of mineral bodies, a very slight variation in their proportionate quantities or combination serving to produce entirely different minerals, or even the very same elements in the same relative quantities wearing a totally different mineral aspect according to slight differences in the conditions of their original formation. Therefore it is that chemical analyses have always hitherto failed, and it would appear that they must always fail, to detect many important mineral differences.

For instance, a rock containing 72 silica, 11 alumina, 2.8 oxide or protoxide of iron, 1 lime, 1.2 magnesia, 1.2 potash, 2 soda, and 0.4 water, may either be a granite, or a gneiss, protogine, granulite, quartz-porphyry, felsite, petrosilex, pitch-stone, trachyte-porphyry, obsidian, or pearlstone; and if we take a wider margin for the proportion of silica, say from 62 to 72, increasing some of the other ingredients in proportion, then a rock, such as we have described, may be a trachyte, phonolite, or minette, for in all the rocks we have named similar values of their

elementary constituents occur. Again, a rock, containing 49-50 silica, 12 alumina, 5-10 oxide or protoxide of iron, 5 lime, 2-3 magnesia, 1 potash, 2 soda, and 0-1 water might just as well be a dolerite as a basalt, or a nepheline rock, leucite rock, diabase, diorite, gabbro, hypersthenite, melaphyre, or porphyrite, for in like manner those values occur in all these rocks.

On the other hand rocks, the same in mineral composition, may vary in the values of their chemical or elementary ingredients 10, 20, or even 40 per cent.

The mineral character of rocks is therefore now sought to be determined in doubtful cases by microscopic rather than chemical analysis, or by tracing the different stages of a rock's transition from a compact into a distinctly composite state; for many rocks (as we shall later have occasion to show) are found to pass by gradual stages from an apparently homogeneous mass into states where their mineral ingredients become distinctly and separately developed so as to be readily recognised.

Whilst chemical analysis was thus found insufficient for determining the mineral character of a rock, it derived a new importance from the igneous theory of the constitution of the primary rocks, when these came to be considered as the products of the consolidation of a general molten mass once the sole material of the earth's structure. The different minerals then came to be regarded as of subordinate importance in inquiring into the origin of rocks, and their differing forms of crystallisation or structure to be regarded but as accidental consequences of slightly different circumstances attending the consolidation of the formerly fused mass.

In this view even the sum of a separate analysis (if it were possible) of all the minerals constituting a rock would fail to present a complete picture of its aggregate chemical character, unless the exact proportionate quantity of each mineral could be also ascertained, which is practically impossible, although it has been sometimes roughly attempted.

These considerations led to the present mode of analysis, which is now usually adopted in the case of all rocks indiscriminately, whether compact or granular, homogeneous or distinctly composite. This is what is termed

G