Ichthyology (Gr. a fish, and discourse), the branch of zoology which treats of fishes, the lowest of the great divisions of the vertebrate animals. The class of fishes cannot be said to have been arranged in a strictly natural manner by any systematise, and such an arrangement is impossible until their external and internal structure and embryonic development are better understood; and until zoologists are better agreed as to what constitutes family, ordinal, generic, and specific characters, little harmony of arrangement can be expected. Most classifications of fishes up to the time of Cuvier (including his) were based on the organs of locomotion and the external integument; after him appeared the anatomical arrangement by J. Muller. The older systems were very imperfect from the ignorance of fossil forms, which supply many links otherwise wanting in the chain of ichthyological characters. Aristotle, in the 4th century B. C, first reduced ichthyology, as he did the other branches of zoology, to scientific form; he was well acquainted with the structure and external characters of fishes, which he distinguishes from cetaceans, laying special stress upon the organs of respiration and locomotion and the scaly covering; he gives the names of 117 species, entering into interesting details on their habits.
The system of compilation without observation prevailed until the middle of the 16th century, when Belon, Rondelet, and Salviani laid the foundations of modern ichthyology! Belon gives rude figures of 110 species, Salviani excellent engravings on copper of 99, and Rondelet woodcuts of 234 species, in all three mostly fishes of the Mediterranean. Gesner in the same century borrowed the descriptions of the last mentioned authors, and added some of his own, in his Historic/, Animalium (1551-'6), all arranged in alphabetical order without any attempt at method, embracing however many foreign fishes. Ray and his pupil Willughby, English naturalists of the 17th century, in their Historia Piscium (1686), gave the first attempt at a natural classification of fishes, founded upon the consistence of the skeleton, the form, the teeth, presence or absence of ventral fins, number of dorsals, and character of the fin rays. They divided fishes into cartilaginous and osseous; though their genera are not well defined, the species are so well described that it is generally easy to refer them to their proper place in subsequent systems; the whole number of species is 420. The second volume consists of well executed, tolerably accurate plates.
This work forms an epoch in the history of ichthyology, which from this time began to assume a methodical arrangement. Passing over Plumier, Ruysch, Kampfer, Sloane, Catesby, and many scientific voyagers of this period, we come to Artedi in the first third of the 18th century. This Swedish naturalist completed the scientific classification of fishes commenced by "Willughby and Ray, defining genera and giving them appropriate names. In his Philosophia he divides the class into four orders, founded on the consistence of the skeleton, the branchial coverings, and the nature of the fin rays, as follows: 1, malacopterygians; 2, acanthopterygians; 3, branchiostegous fishes ; and 4, chondropterygians (sharks, rays, and sturgeons). He made a fifth, including cetaceans, which is inadmissible, and the third is badly characterized; the three others are to a certain degree natural. In his Genera Piscium he gives names and distinctive characters of 45 genera, founded on the number of branchiostegous rays (of which he was the first to see the value), on the position and number of the fins, on the parts supplied with teeth, on the form of the scales, and on the shape of the stomach and caecal appendages; most of these genera stand at the present day.
In his Synonymia Piscium he gives the synonymy of 274 species; his works were published after his death by Linnaeus, his early friend, at Leyden, in 1738. - Linnaeus, in the first edition of the Systema Naturae (1735), followed Artedi; but in the next (1740) he began to give the number of the fin rays, a method of distinguishing since found of great value. In his 10th edition (1758) he trusted to his own knowledge, creating a new system, defining genera more clearly, and using a scientific nomenclature; the most important change was in removing cetaceans from the class of fishes, in which since the time of Aristotle they had been placed, and in uniting them with viviparous quadrupeds in the class mammalia. Brisson, in 1756, had already separated them from fishes. Linnaeus, however, committed the error of placing the chondropterygians among reptiles, under the title of amphibia nantes, to which in the 12th edition (1766) he added the bran-chiostegi of Artedi (ostracion, lophius, tetro-dous, &c). He also suppressed the division of fishes according to the nature of the fin rays, and substituted one founded on the presence or absence of the ventral fins and their position in reference to the pectorals, a method which violates many of the true relations of these animals.
Though Linnaeus neglected some of the genera of his contemporaries, and distributed his orders in an unnatural manner, describing only 480 species, his precision of definition and the excellence of his binary nomenclature were of great advantage to the progress of ichthyology, and his division into apodes, jugulares, thoracici and abdominales for a long time held its place in the science. Linnaeus gave an impetus to the study of natural history, which resulted in making it interesting to all classes, and in inspiring princes with a desire to extend its domain; national expeditions were fitted out by England, France, Denmark, and Russia, which came back laden with treasures of the deep for naturalists; among the workers in this great field we can only mention the names of Commerson, Son-nerat, Pennant, Banks, Solander, the Forsters, Forskal, Steller, Otho Fabricius, O. F. Muller, and Thunberg; the scientific journals teemed with descriptions of new species of fishes from all parts of the globe. - The next great contributor to ichthyology was the German naturalist Bloch, whose celebrated work on the "Natural History of Fishes" consists of two parts essentially distinct; the first, the "Economic History of the Fishes of Germany," appeared at Berlin in 1782-'4, in 3 vols. 4to, with 108 folio plates; the second, the "History of Foreign Fishes," in 1785-'95, in 9 vols. 4to, with 324 folio plates; both were translated into French in a few years after each volume appeared.
Of German fishes he describes 115 species, mostly observed by himself. As he was little conversant with the anatomy of fishes, some of his genera are based on purely artificial characters, while others are remarkably correct. He follows the method of Linnaeus, bringing back the amphibia ncntes, however, into the class of fishes, and dividing them, with Artedi, into branchiostegi and chondropte-rygii. - Comparative anatomy had made considerable progress toward the end of the 18th century, when Lacepede began his researches (1798-1803). He divides the class into cartilaginous and osseous fishes, in each of which subclasses he makes four divisions: 1, with neither opercula nor branchial membrane; 2, without opercula, and with a branchial membrane; 3, with opercula and without branchial membrane; and 4, with both opercula and branchial membrane. In each of the eight divisions he adopts the orders of apodes, jugu-lares, thoracici, and abdominales, according to the absence of ventrals, or their position on the throat, thorax, or abdomen. The natural history of fishes in Sonnini's Buffon (1803-'4) is essentially a copy of Lacepede without acknowledgment. These works of Bloch and Lacepede supplied the principal foundation for most subsequent systems.
The classification of M. Dumeril, in his Zoologie analytique (1806), resembles that of Lacepede, inasmuch as it lays stress upon the supposed absence of opercula and branchial rays and the position of the ventrals. Pallas, in the third volume of the Zoagraphia Russo-Asiatica (1811), gives a list of 240 species, distributed into 38/genera, with the exception of three taken from Linnaeus; he makes two orders, spiraculata or chondro-pterygians, and branchiata, forming with reptiles (pulmonata) the class monocardia (single-hearted or cold-blooded animals). In 1815 Rafinesque published a second ichthyological system in his "Analysis- of Nature, or Tableau of the Universe " (1 vol. 8vo, Palermo); though containing many errors, this system is valuable for several true affinities between fishes before and since regarded as widely separated, as for instance that of the polypterus with the sturgeon family. - De Blainville in 1816 (Journal de Physique, vol. lxxxiii.) published a classification in which fishes are divided into gnatho-dontes or osseous and dermodontcs or cartilaginous, the latter distinguished by having teeth adherent only to the skin; the former include the heterodermes or branchiostegi, and the 8quammodermes or common fishes; in the subdivisions the Linnaean character of the position of the ventrals is adopted, and the families are established principally on the form of the body; it does not employ the Lacepedean characters taken from the opercula and branchial rays. - Cuvier in 1817, in his Regne animal, divides fishes into chondropterygian and osseous.
The former contain the families of suckers (lampreys), selachians (sharks and rays), with fixed branchiae, and the sturionians (sturgeons), with free branchiae. In the osseous fishes he suppresses the branchiostegi, forming of a portion of them the order plectognathi, from a peculiar mode of articulation of the jaws, including the families gymnodonts, scle-roderms, and lophobranchs. The remaining osseous fishes he separates into the orders mala-copterygians and acanthopterygians, after Artedi, according as the rays of the dorsal fin are soft or spiny. The soft-rayed order he distributes into families according to the Linnaean method of the position of the ventrals, disregarding entirely characters drawn from the opercula and branchial rays. The spiny-rayed fishes form a single order, with the families taenioids (ribbon fishes), gobioids (blennies and gobies), labroids (bass), percoids (perches, a very extensive family), scomberoids (mackerellike, also numerous), squammipennes (chaeto-dons, &c), and the flute-mouths (fistularia, &c). He thus makes in all 22 families, founded on direct observation and comparison, and not simply compiled from previous authorities.
Goldfuss ("Manual of Zoology"), in 1820, adopted the four orders of Gmelin, giving to them Greek names, and subdividing them into four families, each according to the shape of the head, mouth, or body, or other external character. - Thus far the systems have been little more than repetitions of the combinations of Artedi, Linnaeus, and Lacepede. Comparative and philosophical anatomy began to be studied with zeal from the beginning of the 19th century. Oken, Carus, Geoffroy Saint-Hilaire, Spix, Weber, Van der Hoeven, Meckel, EVerard Home, Hunter, Tiedemann, and others, wrote upon different portions of the structure of fishes, and the results of their studies began to modify ichthyological classifications. Before mentioning the anatomical and embryo-logical systems, the classification adopted in the Histoive naturelle des poissons, by Cuvier and Valenciennes, beginning in 1828 and coming down to 1868, may be alluded to. In this, fishes are divided into osseous and cartilaginous, the latter (or chondropterygians) including the families sturionians, plagiostomes, and cyclostomes.
The osseous fishes have the branchiaea pectinated or laminated, with the exception of the lophobranchs, which have them in the form of tufts; all the acanthopterygians have the upper jaw free, including 13 families, and all the malacopterygians except the scleroderms, gymnodonts, and lophobranchs; the malacopterygians are divided into abdominals, subbrachians, and apodes. Cuvier had very abundant materials at his command, embracing the collections of Peron, and those of the expeditions under Baudin, Freycinet, Du-perrey, Dumont d'Urville, and other French naval officers. - Oken, in his "Physiophiloso-phy " (Ray society edition), calls the class glos-sozoa, as those animals in which a true tongue makes its appearance for the first time, and os-teozoa, because in them also the bony system first appears. He makes four divisions, the cartilaginous and apodal jugulares, thoracici, and abdominales, the first two having an irregular and the last two a regular body. Among the systems based upon that of Cuvier are those of Bonaparte, Swainson, Straus-Durck-heim, and Rymer Jones. The classification of C. L. Bonaparte (Rome, 1831) comprised the orders: L, acanthopterygii, with 17 families; II., malacopterygii, with 12 families; III., plecto-gnathi, with 2 families; and IV., cartilaginei, with 5 families; including in all nearly 3,600 species.
The principal improvement on the system of Cuvier is in the series in which the genera are placed. Swainson (" Monocardian Animals," in Lardner's " Cyclopaedia," 1838-'9), true to his quinary system, divides fishes into the five orders acanthopteryges, malacopteryges, cartilagines, plectognathes, and apodes. Straus-Durckheim (Traite (d'anatomie comparative, Paris, 1843) adopts the eight orders of Cuvier, but subdivides the chondropterygians with fixed branchiae into three orders, and separates the sharks as the order selaciens, the rays as the order batoides, and the cyclostomes as the order galexiens (from Gr. lamprey), the term cyclostoma having been used for a gasteropod mollusk; he thus makes ten orders. Rymer Jones (in the article "Pisces " in the "Cyclopedia of Anatomy and Physiology," 1847) adopts a modification of Cuvier's system. He makes three divisions: I, chondropterygii or cartilaginous fishes; II., osteopterygii or bony fishes; III., dermapterygii, with skeleton cartilaginous or membranous, and with orders cyclostomata (lampreys) and Iranchiostomata. - About 1830 Prof. Agassiz, principally from the study of fossil fishes, established a classification based on the characters of the scales, as follows: order 1, placoids, corresponding to the cartilaginous fishes of authors, but excluding the sturgeons; 2, ganoids, including the sturgeons, and especially the fossil genera with enamelled scales; 3, ctenoids, comprising bony fishes with scales pectinated on the posterior border, and corresponding generally to the acanthopterygians of Artedi, exclusive of the scomberoids, labroids, and pleuronectes; 4, cycloids, including the malacopterygians with the above exceptions, and exclusive of the blennioids and lophioids.
This system, soon abandoned as an exclusive one by its author from its placing too much stress on external characters, was valuable as connecting in a continuous series living and fossil fishes, and led to the discovery of many important relations between the scales and the internal organs.- The system of Johannes Muller, as given in the Berlin "Transactions" for 1844, derives its characters from anatomical structure, leading often to combinations without regard to zoological differences. He makes six subclasses: I, dipnoi ; II., teleostci ; III., go-noidei; IV., elasmo-branchii or selachii; V., marsipolranchii or cyclostomi; VI., lepto-cardii. Siebold and Stannius adopt this classification in their "Comparative Anatomy;" and a slight modification of it may be found in the third volume of the "Organic Nature " in Orr's " Circle of Sciences," 1855. Owen's classification, mentioned below, and adopted by Sir John Richardson in the article "Ichthyology" of the " Encyclopaedia Britannica," is based partly on that of Muller. - Vogt, in his Zoologische Briefe (1851), divides fishes into the orders leptocardia, cyclostomata, selachia, ganoidea, and teleostia. Van Beneden's em-bryological system (1855) is nearly the same; his orders are plagiostomi, ganoidei, teleostci, cyclostomi, and leptocardii.
Van der Hoeven's classification (as given in the English translation of his •" Handbook of Zoology," 1858) makes fishes the 14th class of the animal kingdom, and divides them into 5 sections, with 11 orders and 46 families. The sections are der-mopterygii, chondropterygii, ganolepidoti, osteopterygii, and protopteri. Milne-Edwards, in his Cours elementaire d'histoire naturelle (1855), divides fishes into osseous and cartilaginous; the former includes the orders acanthopterygii, abdominales, subhrachii, apodes, lophhobranchii, and plectognathi; and the latter, the orders sturiones, selachii, and cyclostomi. - Owen's classification in his " Lectures on Comparative Anatomy" (1855) made the orders dermopteri, malacopteri, pharyngognathi, anacanthini, acanthopteri, plectognathi, lopho-branchii, ganoidei, protopteri, holoccphali, and plagiostomi (sharks and rays). His classification of 1866 is somewhat different, as follows: In the division ho?matocrya, or cold-blooded animals, including fishes, batrachians, and reptiles, in the fishes he makes subclasses: 1, dermopteri, with orders cirrostomi (lancelet) and cyclostomi (lampreys); 2, teleostomi, with orders malacopteri (soft-rayed fishes), anacanthini (cod), acanthopteri (spiny-rayed fishes), plectognathi (ostraceans), lophobranchii (pipe fish), and ganoidei; 3, plagiostomi, with orders holocephali (chimaera), plagiostomi (sharks and rays), and protoptiri (lepidosiren). - Prof. Huxley places fishes in the lowest of his three great divisions of vertebrates, the ichthyopsi-da, including also the batrachians, from the possession of gills, either permanent or temporary; hence he calls them also branchiate vertebrates.
He divides the class pisces into six orders: 1, pharyngobranchii (amphioxus); 2, marsipolranchii (lampreys and hags); 3, teleostci, ordinary fishes; 4, ganoidei; 5, elasmo-branchii, sharks and rays; 6, dipnoi (lepidosiren). - A new classification was published by Prof. Agassiz in his " Essay on Classification," p. 187 (1857), the result of the systems of Cuvier and Muller and of his own scale method, with additional light from his extensive anatomical and embryological researches. He divides the old class of fishes into four; his 1st and lowest class is myzonts, with two orders, myxinoids and cyclostomes; 2d, fishes proper, with two orders, ctenoids and cycloids; 3d, ganoids, with three orders, coelacanths, acipenseroids, and sauroids, and doubtful, the siluroids, plectognaths, and lophobranehs; he was then doubtful whether this class should be separated from ordinary fishes; and 4th, selachians, with three orders, chimaerae, galeodes, and batides. These classes he regards as equivalent to amphibians, reptiles, birds, and mammals. - The following have been the principal cultivators of this science in America: Dr. Samuel L. Mitchill published in vol. i. of the " Transactions of the Literary and Philosophical Society of New York" (1815) a history of 149 species of New York fishes, with many illustrations; he adopts the Linnaean system; other descriptions of his species are in the "Proceedings of the Philadelphia Academy" and in the "Annals of the Lyceum of Natural History of New York." Lesueur has described and exactly figured many species in the Philadelphia academy's "Proceedings." Rafinesque published in the same work, and in his Ichthyologia Ohiensis (1820), descriptions of many species which had escaped his predecessors.
Dr. Kirtland (1838) described the fishes of the Ohio river, and Dr. Holbrook several years later those of South Carolina. Dr. De Kay in 1842, in his " Zoology of New York," divides fishes into bony and cartilaginous, the former having the sections: 1, pec-tinibranchii, with spiny-rayed and soft-rayed abdominal, subbrachial, and apodal orders; 2, lophobranchii, and 3, plectognathi; the latter include the sections eleutheropomi, plagiosto-mi, and cyclostomi. Dr. D. H. Storer, in his "Report on the Fishes of Massachusetts" (1839), and in the illustrated edition of the same in the "Memoirs of the American Academy " (1855-'60), and also in his " Synopsis of the Fishes of North America" (" Memoirs of the American Academy," vol. ii., 1846), follows the arrangement of Cuvier. These works are of great value to the student of North American ichthyology. The Wilkes, North Pacific, and Japan expeditions sent out by the United States government, and the various explorations by land for the survey of the Mexican boundary, the Pacific railroad route, and military and civil roads, have added largely to the materials, both foreign and native, at the disposition of American ichthyologists; these have been worked up principally by Messrs. Baird and Girard of the Smithsonian institution, where the collections are deposited.
The results are published in the government reports on the naval expeditions, in vol. x. of the "Pacific Railroad Reports," in vol. ii. of the " Mexican Boundary Survey," and in the publications of the Philadelphia academy. - The disposition to make new genera and subdivide old ones is carried to a puzzling extreme in ichthyology as well as in other departments of zoology; and the prevalent system of placing the name of the genus maker after the species, by whomsoever and whenever described, offers a premium for naturalists to make the greatest number possible of new genera. In getting rid of the too great condensation of Linnaeus, naturalists have fallen into the worse extreme of too extensive subdivision. For details on the structure and physiology of fishes, see Fishes. - Fossil Ichthyology. Fishes are by far the most numerous of the vertebrates found in the strata of the earth, extending from the Silurian epoch to the tertiary; their number, excellent state of preservation, and remarkable forms, render fossil fishes of great interest in explaining the changes of our planet's surface, and in completing the chain of ichthyic relations.
The classic work on fossil fishes is the Recherches sur les poissons fossiles, by Prof. Agassiz (1833-43); in this magnificent work about 1,000 species are described, with accurate and elegant illustrations, the result of his examinations of more than 20,000 specimens in the cabinets of Europe. He divides fossil fishes into the four orders of ganoids, placoids, ctenoids, and cycloids, according to the structure and form of the scales, these portions of the external skeleton being generally well preserved; the orders he divides into families according to the structure and position of the fins, the form of the bones of the head and of the teeth, and the structure of the gill covers and of the spinous fin rays. His classification is as follows: order I., ganoidei, characterized by osseous plates covered with enamel (see Ganoids) ; order II., placoidei, with tabular scales, like sharks and rays; order III., ctenoidei, having many living representatives, with scales serrated on their posterior margins; order IV., cycloidei, with elliptical or circular scales without serrations.
The first order is most abundant from the old red sandstone to the chalk formation; the second extends from the Silurian through the tertiary epochs; the last two are not found anterior to the chalk, from which they extend through the tertiary strata. For details on fossil fishes, see the geological works of Hugh Miller.