the various substances forming the shell or crust of this vast globe? Impressed with this idea, Patrin has pronounced that the best arrangement is that which is the most lax, because pretended precision would in itself be a radical error: for nature is not regular, but free; and it becomes part of the perfection of a system to partake of that freedom. To expect, therefore, mathematical exactness, or metaphysical acuteness, in the arrangement and nomenclature of natural history, would be foreign to the very nature of the science itself; and if even the most precise and mathematical terms could be found, they would be improper in mineralogy, where the substances themselves are inaccurate, and all the divisions are mutually intermingled, and graduate into each other. In the Substantial Domains even complex rocks, as granite, &c. are equally simple with some substances regarded as homogenous; and smaragdite, for example, will present as numerous ingredients. But in the Substantial Domains the Modes are variations of the same substance, and naturally follow each other; while the Nomes are compounds wholly different in themselves, and connect detached substances in an arrangement totally distinct. In the former case the terms themselves may be regarded as definitions, which is a great advantage in any science; while the Nomes must, from the very nature of the subject, be considered as arbitrary divisions for the sake of memory. In this point of view a system may be compared to a cabinet; and if each substance can find its proper drawer and place, the object of utility and clearness are answered. But at the same time every system, even the Newtonian, has its anomalies.

In this, as in the former part, it became a chief object to Nomenclature. increase the nomenclature, the poverty of which has long been regretted by Saussure, and other able authors. Buffon presents some useful observations on this topic. "Men have begun with giving different names to things which have appeared to them clearly distinct; and at the same time they have formed general denominations for objects which seemed to resemble each other. Among savages, and in all new lan

Hauy's deviations.

INTRODUCTION.

guages, the names are almost always general, that is to say, vague expressions for objects of the same similitude, however distinct. An oak, a beech, a linden, a yew, a pine, a fir, will all at first be called a tree; then the oak, the beech, the linden, will all be called oaks, till they be distinguished from the others, which will be called pines. But particular names will only be found in an advanced state of society, after comparisons and examinations; and the number has been always increased in proportion as nature is more studied and better known; and the more it is examined and compared, the more abundant will be the proper names and peculiar denominations. But when we are now presented with general terms, that is genera, it is to send us back to the A B C of knowledge, and recall the darkness of the infancy of nations. Ignorance has created genera, science has produced, and always will produce, proper names; and we are never afraid to augment the number of particular denominations when we wish to designate different objects."

This eloquent author was, however, too inimical to systems of nomenclature on the Linnæan plan; and his observations may be considered as chiefly applicable to mineralogy, in which the arbitrary divisions have been so often confounded, as has already been explained in the general introduction to this work. The most severely scientific writer on mineralogy is Haüy, but even he has been obliged repeatedly to change the subdivisions; for in the first class he has genera, in the second only species; in the third there are two orders; in the fourth three orders, and every metal forms a genus. Nay, as already stated, he has changed the very foundation of his plan, having formally abandoned the integrant molecule, which, as he supposed, constituted the species, for the primitive form, as he confesses that he was often deceived by the integrant molecule *. This molecule was the invention of

His argument that crystals resemble the flowers of plants, as a criterion of species, is not just, the crystals being often different from the substance, quartz in limestone, barytes in granite, &c. &c.

the ingenious crystallogist Romé de Lisle; and formed the foundation of the singular production of Dolomieu on the Mineralogical Species, in which he goes so far as to assert that this species can propagate itself! This nugatory propo-, sition seems merely to have been advanced, because he allows that without this quality no species in natural history can exist. Let it not be imagined that such observations, extorted merely by the impulse of truth, are intended to violate the respect due to those great writers in other departments of the science, which is sufficiently wide for the developement of various talents; and though the eagle requires a whole province of rocks for his immediate domain, there is in this science ample space for invention and ability, without enmity and without envy.

It is hoped that the nature of the several domains contained in this volume will be found to be sufficiently illustrated by the observations at the head of these divisions. One of the most important, in every point of view, is the Volcanic, an object of ludicrous neglect and contempt to the German mineralogists, whose confined ideas have been the more implicitly followed, because the Germans are the fathers of modern mineralogy. It will here be found to be treated with the details, and it is hoped with the accuracy, which the subject deserved, not only from its own importance, and contradistinction from all the other domains, but on account of the infinite contestations which have arisen on this topic among the most eminent writers in the science. Diffident, however, of his own ideas, it gave the author singular satisfaction to find them confirmed by those of the first chemist of this or any age, as may be judged by the following extract from one of our weekly journals*.

Observer, June 2d, 1811.

Volcanic

rocks.

门

INTRODUCTION,

Dr. Davy's "In the concluding lecture, Dr. Davy stated, that the observations. emission of lava from volcanoes was one of the principal operations by which nature supplied the waste of rocks, and the destruction of the land noticed in his former lectures. The agency of volcanoes in the production of islands, and the increase of continents, is more extensive, than those who reside at a distance from their influence are disposed to admit. Proofs of this may be traced in the islands and shores of the Mediterranean, in the continent of America, and in Asia, and in other parts of the globe. Nearly the whole of Sicily, and the southern parts of Italy and France, offer evidence of their volcanic origin; and Rome, which has by ancient writers been proudly styled the "Eternal City," is built on the crater of an extinct volcano. The phenomena attending the eruption of volcanoes were described from Hamilton, Dolomieu, Spalanzani, and others, who had been present during the eruptions of Etna and Vesuvius.

"The convulsion of the solid ground, the lofty columns of flame, smoke, and vapour, the tremendous explosions, the torrents of rain, and the thunder and lightning, which accompany the eruption of lava, all indicate that the immediate cause is the expansion of steam and hydrogen gas, which inflames when in contact with the atmosphere. The doctrine of a central fire was unsupported by proof or analogy: did such a fire exist its effects must be felt at the surface, even if it had to pass through the most imperfect conductors of heat.

"To ascertain the cause which produced the expansion of vapour, and the other phenomena of volcanoes, we must examine the products of these august operations of Natural Chemistry. If we observe. a fire at a distance, and are able to collect its products, we may thence determine the nature of the substances which have been in a state of combustion. The products of volcanoes are hydrogen gas, vapour, and lava, of which lava is a compound of the earths, the alkalies, and the oxyd of iron.-In his former lectures, Dr. Dayy ob

served, that he had stated the discovery of the metallic nature of the earths and the alkalies, and that the metals obtained from them were in a high degree inflammable when they came in contact with water.-Dr. Davy further observed, I that previous to the eruption of volcanoes, the lakes and springs in their neighbourhood were known to have suddenly disappeared; and all the volcanoes which are in activity are in the neighbourhood of the sea, or of large lakes*. Now if we admit that these earths exist under the surface, in a metallic state, the access of water to them would occasion their combustion. The oxygen would be absorbed, and an immense volume of hydrogen gas would be produced, which is always found to follow the eruption of flames. This explanation of the cause of volcanoes may be considered as a reasonable influence from the discovery of the metallic nature of the earths; and if we admit the operation of electrical agency in the globe, we shall have a cause operating by which the earths may be restored to their metallic form. Thus the process of renovation and decay will be constantly balancing each other, and nature be preserved in a state of eternal youth. The appearance of the Aurora Borealis and the Aurora Australis, render it probable that the poles are in two different states of electricity, and that a constant circulation of electric power is taking place.

"Though new land and soil may thus be formed, Dr. Davy said he was not inclined to admit that the primary and secondary rocks were thus produced. The crystals they contain are different from those ever found in lava. The experiments of Sir James Hall, which had been thought to establish the volcanic nature of basalt, he considered as defective. In basalt, hornblende and felspar are distinctly crystallised; but the fused basalt which had slowly cooled, though it had the form of basaltic prisms, did not contain hornblende or felspar in distinct crystals.

The Andes are from 80 to 100 miles distant from the sea, and are only influenced by subterranean lakes. P.