Alexander von Humboldt, who has enjoyed the rare privilege of being able to assist so many men of talent,—was from the first the devoted friend of Agassiz, and it was his patronage that enabled our naturalist to commence in 1833, so soon after his arrival in Switzerland, the publication of his great work on Fossil Fishes; which he dedicated to Humboldt, and of which we intend to say a few words, as of all his works this made the greatest sensation, and it is this that obtained for him the eminent rank which he now holds in the scientific world.

This work consists of five volumes, with an atlas of about four hundred folio plates, and comprises descriptions and figures of nearly a thousand species of fossil fishes. All the specimens are represented of the natural size, with the colors of the bed from which they were taken. It was impossible that so many new species should be made known without rendering many alterations necessary in the science of Ichthyology; new types were established, and the affinities of various groups and families to each other more clearly shown. Moreover, Agassiz did not confine himself to establishing a vast number of species, genera, and even families. Beside this he founded an entirely new classification, based principally on the importance of the fossil fishes.

Cuvier makes two general divisions among Fishes; the Osseous and the Cartilaginous fishes. Agassiz also separates the Osseous fishes from the Cartilaginous, of which he makes his first order, that of the Placoïdians; but he divides the Osseous fishes again into three other equally important orders; so that the class of Fishes is divided into four orders; namely, 1. The Placoïdians; 2. The Ganoïdians; 3. The Ctenoïdians; and 4. The Cycloïdians. This classification is not founded on the skeleton, like that of Cuvier, but on the nature of the outward integuments, the scales. Agassiz starts with the principle that the outward covering of fishes is the reflex of their internal organization. With this principle he examines the different families of the class of Fishes, with respect to their scales, and finds in the conformation of the external integument a variety of characters, on which he founds his classification. As to this, it is to be remarked at the outset, that all the Osseous fishes, with the exception of a few genera, are furnished with horny scales; while the skin of the Cartilaginous fishes is covered with plates or spines, of a peculiar form, known under the names of shagreen, &c. The scales of the

Osseous fishes are constructed on a totally different plan, and the differences are so marked, that Mr. Agassiz considered them a sufficient foundation for his three orders of Cycloïdians, Ctenoïdians and Ganoïdians. The two former, which comprise almost all the Osseous fishes of the present epoch, both have horny scales; but they differ in this, that the Ctenoïdians have the posterior edge of the scales indented, while in the Cycloïdians this border is entire. He seeks to prove that this distinction, apparently insignificant, is, in truth, founded in nature, being the expression of a fundamental character which reveals itself equally in other parts of the body. Thus, fishes having indented or pectinated scales, have generally prickles on the head, the opercula, and various parts of their body; while the others-the Cycloïdians-are smooth, and without defence. Mr. Agassiz considers the Perch, with the analogous species, as the type of his order of Ctenoïdians; and the family of the Carp, Salmon, Pike, &c., as typical of the Cycloïdians. This division corresponds, therefore, to a certain extent, with Cuvier's division of fishes into Acanthopterygians and Malacopterygians.

The second order, that of the Ganoïdians, seems to have a yet more satisfactory foundation. There have been found in the Nile and in the rivers of North America, two fishes which have always puzzled the ichthyologists; that of the Nile is known under the name of Bichir (Polypterus Bichir); the other, which is found in America, is called the Gar-pike (Lepidosteus), having some resemblance to the Pike. Both these fishes are furnished with scales of very peculiar form and structure. Instead of being arranged in the manner of rooftiles, -as in most fishes,-they are placed simply side by side, the surface being covered with a coat of enamel, making a very solid cuirass. On examining these fishes in an anatomical point of view, Mr. Agassiz found that the skeleton presented no less striking differences than the scales and the soft parts of the body. Nevertheless, it seemed hazardous to separate them altogether from the other great families; and particularly when the smallness of their number was considered, it seemed contrary to all method to place them in the same rank with the Placoïdians on one side, and the Osseous fishes on the other. But the procedure, though not authorized by the study of the living fishes, was justified by an examination of fossil species. Here is displayed a whole ichthyological fauna, having the characters neither of the Osseous

nor of the Cartilaginous fishes, but altogether analogous to the Bichir and the Lepidosteus. So that these two genera, apparently mere exceptions in the present creation, in reality constitute a type by themselves, which, though not numerous at present, is, nevertheless, the expression of an entire order of things. Associating with these fishes the numerous fossil species whose scales have the same structure, Mr. Agassiz made his division of Ganoïdians, which already contains many hundred species, and promises to become still larger, since it predominates in all the formations anterior to the chalk. Mr. Agassiz recognizes several distinct families of this order; the two principal ones are the Sauroïdians, to which the Lepidosteus and the Bichir belong; and the Lepidoïdians, which were inoffensive and probably omnivorous fishes, somewhat resembling the Carp in appearance, but having no representatives in the present creation.

These researches among the fossils had not a geological interest alone. The numerous examinations that Mr. Agassiz was obliged to make, in order to establish in all points the analogy of extinct species with living types, revealed to him anatomical relations of great interest, which had been hitherto passed over. He thus discovered the important fact, not before made known, that there exists a remarkable parallelism between the development of the individual, and the development of the whole class in the series of ages. In the early stages of embryonic life, the vertebral column does not exist. In place of it there is found, in the embryo, a gelatinous mass, called the dorsal cord. Around this cord (which remains for a longer or shorter time in all fishes,) are formed the vertebræ, as bony rings. These rings gradually increase, and encroach more and more upon the dorsal cord, which, in most fishes, at last disappears. In some types, however, for example, in the Sturgeon, it remains during the whole life; so that this fish has no vertebræ, and the apophyses rest immediately on the dorsal cord. Now, Agassiz shows us that this is the case with all the fishes of former epochs. They all have distinct spinous apophyses, often very strong and completely ossified, but they show no trace of separate vertebræ ; whence he concludes, that these organs were wanting, and that the dorsal cord continued throughout life, as in the Sturgeon. As o the relative superiority of living types, also, embryology reveals to us a wonderful parallelism. There is no fish, however imperfect, whose organization does not corre

spond to some phase in the life of more perfect types. Take, for example, the Lamprey, or that still more imperfect fish known under the name of Amphioxus, or Branchiostoma, which Pallas placed among the Snails, from its great dissimilarity to ordinary fishes. The former has, in place of the cranium, only a cartilage corresponding to the base of the skull; and the latter is deprived even of this, and the dorsal cord extends to the end of the snout. The first has a single fin, more or less divided; in the other, the fin extends along the whole body. Finally, neither has jaws, properly so called. Now, the most perfect of our fishes, such, for example, as the Salmon, are all, at one period of their life, at the same point of development, but with them it is a transient state, a stage of growth; whilst in the others it is the permanent condition.

These views have a high philosophical bearing, particularly in their application to other classes of the animal kingdom. It is in accordance with them that Agassiz determined the rank to be assigned to the various families of fishes, according to their organization.

It is to Geology, nevertheless, that the greatest profit is derived from these discoveries. In comparing together the fishes found in various formations, Agassiz from the first had also thrown new light on the relative age of these formations. Thus, to cite but a single example, he was enabled by the study of the fishes of the slate of Glaris, to demonstrate that this deposit, which had previously been considered as belonging to the most ancient sedimentary rocks, the grauwacké, is much more recent, and forms a part of the cretaceous group. Another and more general result of his labors was the discovery, that not only are all the fossil species different from those now living, but also, that from one formation to another, the species are equally distinct. And this diversity, according to him, is not confined to the larger formations, but exists equally between the various stages of the same formation. Thus he recognizes no species as common to the lias and the upper Jura limestone; to the upper and lower cretaceous deposits; to the ancient and recent strata of the tertiary formations, &c. The necessary deduction is, that the whole creation has been renewed at different epochs, by a direct intervention of the Creator. Agassiz, however, did not stop here, but pushed his conclusions still further. From the fact that certain basins, like certain regions of the earth's surface, are inhabited by species peculiar to them, not found elsewhere

in deposits of the same age, he inferred that each creation. was local, that is to say, that species were created in the localities they inhabit, and that to each was assigned a limit, which it does not pass so long as it remains in its natural condition. Man alone, and those few species that are associated with him, are exceptions to this general law. And as the migrations of even these species takes place under the direct influence of man, we may conclude that they were unknown to former epochs.

These considerations, with others not less important, concerning the relation which this localization bears to the temperature and degree of elevation of continents at different epochs, suggested to Agassiz some general reflections, with which he closes his chapter on Classification, and which we transcribe, as showing the spirit in which this work is written. "Such facts," says he, "loudly proclaim principles which science has hitherto left untouched, but which the researches of paleontology urge upon the observer, with an ever increasing force: those, I mean, that respect the relation of the Creator to the universe. We see phenomena closely connected in the order of succession, yet without any sufficient cause within themselves for the connection; an infinite diversity of species, without any material bond of union, so grouped as to present the most admirable progressive development, in which our own species is involved. Have we not here the most incontestable proofs of the existence of a Superior Intelligence, whose power alone has been able to establish such an order of things? The methods of scientific investigation, however, are of such strictness, that what seems to our feelings a matter of course, we cannot admit, unless supported by numerous and well-established facts; on this account, I have delayed expressing my convictions on this subject, until the last moment; not that I have wished to avoid the discussions which the announcement of such results must necessarily excite, but that I have been desirous not to provoke them before establishing for these results a purely scientific foundation, and supporting them by rigid demonstrations, rather than by a profession of faith. An acquaintance with more than fifteen hundred species of fossil fishes, has taught me that species do not pass insensibly into each other, but that they appear and disappear unexpectedly, without showing any immediate connection with those preceding them. For I do not think that any one can seriously affirm that the nu