Page images

ly stated: But as we do not consider that the excellent author has examined this subject with much care or affection, we shall not proceed with our extract, but rather endeavour to lay before our readers our own views of this important question.

A classification of rocks on a Mineralogical principle must unquestionably be considered, in one sense at least, as a natu ral arrangement; associating those combinations of minerals which are actually found in nature, just as in the organized world we associate certain combinations of forms. It may also be contrived in such a manner as to distinguish the simple from the compound rocks, and thus to refer the former to their proper places in the system of mineralogy; while it is more exclusively occupied in the classification of the latter.

As, in nature, the same compounds occur in different Geological positions, it follows, that any arrangement on a Geological principle must involve repetitions, which cannot occur in one founded on mineralogical characters. It is also evident, that from the present imperfect state of the science, much must be assumed or conjectured respecting the general order of succession among rocks, and the analogies which they bear to each other; so that, to a certain degree, every geological arrangement must be hypothetical, while no hypothesis is involved in a mineralogical one.

Were a sufficiently extensive nomenclature provided and received, it is evident that a mineralogical arrangement would furnish a name for every rock; and that it would thus be enabled accurately to limit the same term to the same compound, whatever might be its position in nature. To all which advan

tages it may lastly be added, that the knowledge of rocks classified on this principle, demands no geological investigations; and is open to every one who has acquired the previous knowledge of minerals in their simple state.

Now we most readily admit, that, as far as the mineral history of the species or varieties of rocks is concerned, a mineralogical classification is the best that could be adopted; and that it would, if perfect, materially facilitate the examination or description of a cabinet of specimens. Such a classification could only, however, be considered as part of a system of mineralogy. As yet, all these systems must be considered as artificial; classing, in the manner most convenient for investigation, those substances of which the obscure and involved affinities have as yet prevented all possibility of a natural arrangement. If the simple rocks, therefore, are to be described among the simple minerals, as in the system of Brongniart, the compound, to render the system uniform, should be enumerated in the ca

talogue of accidents to which each mineral is liable with respect to mixture or association. If this be not adopted to its full extent, it is obvious that rocks, objects of such importance, not only in the history of the earth, but in that of the minerals themselves, become separated into two distinct works or catalogues. As it is also not uncommon for a single mass of rock to be compounded in one place and simple in another, it would be necessary to search for it in two distinct catalogues; and possibly, in the imperfect state to which all new arrangements must be subject, in the works of two different authors.

The difficulties to which we have just alluded, which have hitherto prevented us from establishing a natural, and at the same time an useful classification of minerals, interfere even in a greater degree with any attempts to form a natural and an useful arrangement of rocks on a mineralogical basis. The most important characters of these are frequently to a considerable degree independent of the minerals which enter into their composition. Not unfrequently, also, a rock will retain all its most essential qualities, although undergoing considerable mineralogical changes, by the loss of one, or the acquisition of another substance. It must also be evident, that the capricious, and almost endless modes in which the minerals that constitute rocks are intermixed, would lead to an enumeration of species that would confound the student by its excess, or, if curtailed, defeat the object of the contriver.

It is, therefore, an important defect in such a classification, that it bears no necessary relation to some of the most interesting characters of rocks. But it is also difficult to propose any mineralogical method which is uniform and unexceptionable; even were we to overlook the advantages derived from one which should fulfil that condition to which we have just alluded. Let us suppose, for example, that the presence of some particular mineral is made the ground of association. In such a case, it is evident, that substances most widely separated, not only in their natural affinities, but even in their characters as mere specimens, or rocks, may be associated together; since the minerals which enter into their composition are very few, and are repeated under many different combinations throughout a great number of rock species. We need not quote examples to the geological reader of the confusion that would thus ensue, for example, by assuming either quartz, mica, or felspar, as the common bond of a class, or of any inferior division.

If, again, the predominance of some one mineral be made the groundwork of a division, it is easy to see that the same rock might be separated into different classes or subdivisions of

[ocr errors]

an inferior kind. This, for example, would often happen in the case of granite; which may contain in excess any one of the different minerals that enter into its composition.

In the next place, let it be imagined that the texture of a rock is assumed as a common bond of union, and it will be equally apparent, that under the granitic, the porphyritic, or the schistose, cognate substances may be widely separated, and those which are entirely different in other more essential circumstances be associated. The same reasoning applies to any attempt to arrange rocks according to the number of the substances of which they are composed. Thus, if a binary, or a ternary proportion be made the groundwork of any association, similar effects will follow; as it is not unfrequent for the same rock to vary in the number of its ingredients. It would be abundantly easy to illustrate all these objections by a reference to well-known rocks but the enumeration is scarcely required, and would extend these remarks too far. It must, indeed, be obvious, that any mineralogical arrangement, even should it combine all these methods in the most careful manner, must in a great degree be arbitrary; and that it must hold out the shadow, rather than the substance, of a natural method.

If we even imagine such a system to be perfected, it is evident, as we have hinted above, that it would require a very numerous, as well as an appropriate set of terms; and this, as far as it has been executed by the able author who stands first in the work under review, it actually does. Let us consider, then, how this would affect geological descriptions, the principal object for which a knowledge of rocks is required. There is no necessary relation between the composition of a rock and its place in the order of nature; and many varieties of composition, as we have just seen, occur in the same mass of rock,as in the familiar instances of the porphyries and gneiss. Many terms would, therefore, be required in such cases, to describe one geological fact, or one set of connexions; and it would also follow in other instances, that such connexions would appear to be implied where they did not exist; merely in consequence of the terms by which rocks, similar in composition, but different in geological characters, were designated. The circumlocutions and difficulties that would follow, in the first of these cases, and the confusion that would result in the latter, are too obvious to be stated.

There is another objection to a mineralogical arrangement, which appears to us of no small importance. It renders of equal value those rocks which are rare, and in some measure accidental, and those which are of the greatest consequence and

interest, as far as the structure of the earth, or the natural history of these substances, is concerned. The most rare modifications, the most limited varieties, would thus claim as much attention as those which are the most constant and the most common; while differences, which may be of the greatest importance in geological science, but would not excite much attention from their mineralogical composition, may thus pass with much less. notice than they demand. Hornblende schist will furnish the reader with an easy illustration of this remark. In the same

Geological connexions it may be simple, or it may contain a few particles of felspar; but, according to a Mineralogical system, it must be designated by two distinct terms; and, what is worse, the simple rock will not even be found in the arrangement at all, but must be sought for among the simple substances in the system of mineralogy.

If, therefore, in one sense, a mineralogical arrangement of rocks is natural, it is, in a much more important view, unnatural, or artificial; as it disjoins the wider and more interesting affinities by which these substances are connected with each other and with the general structure of the earth. It is, in fact, an artificial system, with the imposing appearance of a natural one; founded on a minute set of appearances, and negligent of the larger features, and the numerous important circumstances of affinity, or difference, which prevail among the objects of its contemplation. Thus it in some measure resembles the artificial arrangements of the ancient botanists as compared with the more philosophical views of the moderns in their establishment of Natural Orders. Mineralogists, indeed, appear in this instance to have been misled by the example of Linnæus, and by the valuable consequences that have resulted from his systems in the organized departments of natural history; forgetful of the important and radical differences by which these departments are distinguished from the peculiar objects of their study.

Were it possible to make any arrangement, however artificial, which should facilitate the study of rocks as constituent parts of the earth's structure, it would form a valuable acquisition to the geologist, as well as to the collector of specimens. But if, in teaching the latter to arrange his cabinet, it misleads the former-it is injurious and not beneficial. In the present state of our knowledge, it appears indeed a vain attempt, as well as an inconvenient and injurious sacrifice to the formalities of an imaginary logic.

In proceeding to consider the comparative advantages and defects of a Geological arrangement of Rocks, we think that the authors under review have not been sufficiently careful in dis

tinguishing between the study of minerals and that of the structure of the earth. The connexion of the former with the latter is doubtless an important part of their history; but the knowledge of minerals can never form a proper basis for the arrangement of the great masses of which our globe is composed. If our sole object in the study of rocks were a knowledge of their mineralogical composition, such an arrangement would doubtless be the best. But the main end of that study is to investigate their proportions, their gradations, their analogies, their mutual dependence or connexion, their order of succession and disposition-in short, their general relations of all kinds, to each other and to that structure which forms the object and business of geology: And this, as it appears to us, can only be attained by a classification founded on a geological basis. Such a classification has, for its foundation, the most extensive affinities and the most important characters of the objects to be arranged; and it thus in some measure resembles a classification of plants, according to their natural orders. While it instructs us in the history of rocks as constituent parts of the earth, it does not exclude their history as mineral compounds: since we are enabled to combine with the former the most minute subdivisions of varieties; and at the same time have it in our power to separate the accidental and unimportant from the constant and essential. It also appears to us that it affords equal facility for reference as a mineralogical arrangement, by the very simple expedient of brief synoptic tables; thus combining their greater and their lesser analogies; their order in nature with their mineralogical affinities.

It cannot however be denied, that the objections to such a system of arrangement are both numerous and weighty; and it is only by comparing these with the advantages now stated that we can be guided in our choice.

The most formidable objection is the imperfect knowledge which we at present possess of the true order of rocks in nature. Whatever system therefore we adopt as the basis of such an arrangement, must confessedly be imperfect. But it may still be such as to be capable of perfection; and it offers a basis not only susceptible of correction, but gradually increasing in correctness; since every step adds something to the mass of facts on which it is founded.

It must also be admitted, that a geological arrangement cannot be logically correct; since it cannot be founded on one simple and consistent principle. While the larger divisions are derived from the general order which rocks hold in nature, the smaller are necessarily founded on mineralogical characters,

« PreviousContinue »