purely basic mineral magnetite may be placed, as lithologically Class 1st.-Felspithic-Quartz, Orthoclase, Oligoclase, Labra- Class 2nd.-Basic-Potash mica, Magnesia mica, Hornblende, The extent to which these minerals enter into the constitution of original rocks will be best seen by repeating here the general view given of the families of rocks, placing at the head of each column the names of the principal constituents. TABLE II. General View of the Mineralogical Constitution of the families of Original Rocks. Basic Rocks. Baseus Rocks Neutral Rocks Siliceous R'ks Silicic Rocks. It will be observed from this table that a certain degree of consistency is observed by the essential minerals in cntering into the constitution of original rocks. Such acid minerals as quartz and orthoclase never occur in the basic rocks; nor, on the other hand, do we find augite or labradorite entering into the composi tion of siliceous granites or trachytes. Towards the basic extreme of chemical composition in rocks, the siliceous minerals diminish or disappear, and, towards the acid extreme, basic minerals act in the same way. This behaviour alone is sufficient to shew that the mineralogical constitution of a rock is not the result of accident, but mainly the consequence of the chemical nature of the plastic magna from which it resulted, an inference which is borne out by the varying composition of the minerals themselves. It will be seen that at the heads of the columns the minerals have been arranged according to the classification already given. Now it would appear, with regard to the members of each of the classes which we have distinguished, that not only do they resemble each other in chemical composition but they seem to replace each other when they enter into the composition of origi nal rocks. That is to say, the increase, in quantity, of one of them in a rock is generally accompanied by a decrease on the part of another member of the class,and generally of that member which most closely approaches the first in chemical composition. This appears to be well borne out by the table,and numerous examples of such substitutions might be cited. Thus hornblende replaces mica in granitite forming syenite; oligoclase replaces orthoclase in the passage from syenite to diorite; and diallage replaces pyroxene in that species of greenstone called gabbro. There are thus formed gradual transitions from one rock species to another in mineralogical constitution as well as chemical composition. In the subjoined table (III) the nature and manner of these transitions arc exhibited. It will be seen that the distinctions already made as to the orders and families of rocks are kept steadily in view while at the same time an attempt is made to give a systematic arrangement of the different species of original rocks and their mutual relations. Basic Rocks. Basous Rocks. III. Table showing the Mineralogical Constitution of the Species of Original Rocks. Basic non-fe'spathic Rocks. Neutral Rocks. Siliceous Rocks. Silicic Rocks. Silicic non-elspathic Rocks. In preparing table III, the same care has been taken as with those already given to introduce no new terms, and to use the various names of the species only in the sense which at present is generally attached to them by petrologists. In a few instances, where such names have hitherto borne a too general or a more or less indefinite meaning, an attempt has been made to confine their application to one species. The name rhyolite is for instance used in a somewhat more restricted sense than that given it by its originator, and the very vague, generally condemned, but still much used or misused, name, melaphyre, is, as applied to a parti cular species, limited to those porphyrite rocks which are neutral in chemical composition and in which crystals of triclinic felspars only are developed. In some other cases, where the same species possessed several synonyms, a slightly different signification has been given to one, and generally the least used of them, in order to make it of use in our system. For instance, curite and felsite have hitherto been synonymous. In our table the latter term is made to indicate the more silicic species of fine grained rocks. Such names of rocks as have been derived from those of minerals have their terminations, in accordance with Dana's suggestion, altered from ile to yte. It will be observed that, in table III, the minerals of the felspathic class only are placed at the head of the vertical columns, while the other essential minerals have been placed under each variety of texture on the left hand side. The cause of this arrangement may here be stated. The felspars, being of very constant occurrence in original rocks, and being frequently difficult to determine, have not been much made use of in distinguishing species until quite recently. For instance, oligoclase very often can only be distinguished from orthoclase by an experienced mineralogist, and only an experienced chemist after a minute analysis, can distinguish between oligoclase, labradorite and anorthite in a compound rock. On the other hand the minerals of the other class possess very well marked physical characters, and the presence of one or other of them was readily detected by the earlier petrologists and made use of by them for characterising different rocks. Thus, mica, hornblende and olivine are very widely apart both as regards form, colour, hardness and fusibility. The only two minerals of the second and third classes which are difficult to distinguish from each other are hornblende and augite, and this is only the case in fine grained compound rocks. By giving prominence to cach of these non-felspathic minerals and placing their names on the horizontal lines of our table, it becomes possible to shew at a glance the rocks which they form with the felspathic minerals named at the heads of the vertical columns, and the manner in which, by gradually replacing each other, they form the different species of original rocks. Thus it will be observed that among the schistose rocks the most basic is diabase schist; that the latter becomes diorite schist when hornblende replaces pyroxene; that the diorite schist, as its oligoclase is replaced by orthoclase, becomes syenite schist, and, as quartz makes its appearence and increases, syenitic gneiss is produced. At the next step in a silicic direction, mic replaces the hornblende, producing common gneiss, then when the mica disappears, granulite results. If, instead of the mica, the orthoclase disappears, mica schist is developed, and when from the latter rock the mica in greater part is withdrawn, it becomes quartz schist. The other varieties of texture, such as the porphyritic and trachytic, each exhibit a similar series of transitions, the most fully developed being the granular order. In the latter it becomes possible, by means of the peculiar arrangement of our table, to shew the mineralogical nature of each of the species of the complicated family of the greenstones. Diorite, gabbro, hyperyte, diabase and protobastyte rock are shewn to be respectively characterised by hornblende, diallage, hypersthene, pyroxene and enstatite in combination with various felspars. The great majority of original rocks contain some variety of felspar, but there are a few species in which that mineral is absent and which are called non-felspathic rocks. In order as far as possible to shew these also in our table, two columns have been added to it, one at each side. The right hand one shews the silicic, and the left hand the basic rocks void of felspar. VI. ACCESSORIAL CONSTITUENTS. Besides the minerals mentioned in the foregoing chapter as the essential constituents of crystalline rocks, there are others of less frequent and only accidental occurrence, which have been called by German lithologists the accessorial constituents. Among these such minerals are not included as are only found in the veins, cavities, or even joints enclosed in rocks. Only those which are found in intimate mechanical union with the essential constituents in the body of the rock itself are regarded as accessorial consti |