CHAPTER XXI

CLASSIFICATION OF FISHES

=Taxonomy.=--Classification, as Dr. Elliott Coues has well said,[147] is a natural function of "the mind which always strives to make orderly disposition of its knowledge and so to discover the reciprocal relations and interdependencies of the things it knows. Classification presupposes that there do exist such relations, according to which we may arrange objects in the manner which facilitates their comprehension, by bringing together what is like and separating what is unlike, and that such relations are the result of fixed inevitable law. It is therefore taxonomy (~taxis~, away; ~nomos~, law) or the rational, lawful disposition of observed facts."

A perfect taxonomy is one which would perfectly express all the facts in the evolution and development of the various forms. It would recognize all the evidence from the three ancestral documents, palæontology, morphology, and ontogeny. It would consider structure and form independently of adaptive or physiological or environmental modifications. It would regard as most important those characters which had existed longest unchanged in the history of the species or type. It would regard as of first rank those characters which appear first in the history of the embryo. It would regard as of minor importance those which had arisen recently in response to natural selection or the forced alteration through pressure of environment, while fundamental alterations as they appear one after another in geologic time would make the basal characters of corresponding groups in taxonomy. In a perfect taxonomy or natural system of classification animals would not be divided into groups nor ranged in linear series. We should imagine series variously and divergently branched, with each group at its earlier or lower end passing insensibly into the main or primitive stock. A very little alteration now and then in some structure is epoch-making, and paves the way through specialization to a new class or order. But each class or order through its lowest types is interlocked with some earlier and otherwise diverging group.

=Defects in Taxonomy.=--A sound system of taxonomy of fishes should be an exact record of the history of their evolution. But in the limitations of book-making, this transcript must be made on a flat page, in linear series, while for centuries and perhaps forever whole chapters must be left vacant and others dotted everywhere with marks of doubt. For science demands that positive assertion should not go where certainty cannot follow. A perfect taxonomy of fishes would be only possible through the study, by some Artedi, Müller, Cuvier, Agassiz, Traquair, Gill, or Woodward, of all the structures of all the fishes which have ever lived. There are many fishes living in the sea which are not yet known to any naturalist, many others are known from one or two specimens, but not yet accessible to students in other continents. Many are known externally from specimens in bottles or drawings in books, but have not been studied thoroughly by any one, and the vast multitude of species have perished in Palæozoic, Mesozoic, and Tertiary seas without leaving a tooth or bone or fin behind them. With all this goes human fallibility, the marring of our records, such as they are, by carelessness, prejudice, dependence, and error. Chief among these defects are the constant mistaking of analogy for homology, and the inability of men to trust their own eyes as against the opinion of the greater men who have had to form their opinions before all evidence was in. Because of these defects, the current system of classification is always changing with each accession of knowledge.

The result is, again to quote from Dr. Coues, "that the natural classification, like the elixir of life or the philosopher's stone, is a goal far distant."

=Analogy and Homology.=--_Analogy_, says Dr. Coues, "is the apparent resemblance between things really unlike--as the wing of a bird and the wing of a butterfly, as the lungs of a bird and the gills of a fish. _Homology_ is the real resemblance, or true relation between things, however different they may appear to be--as the wing of a bird and the foreleg of a horse, the lungs of a bird and the swim-bladder of a fish. The former commonly rests upon mere functional, i.e. physiological, modifications; the latter is grounded upon structural, i.e., morphological, identity or unity. Analogy is the correlative of physiology, homology of morphology; but the two may be coincident, as when structures identical in morphology are used for the same purposes, and are therefore physiologically identical. Physiological diversity of structure is incessant, and continually interferes with morphological identity of structure, to obscure or obliterate the indications of affinity the latter would otherwise express clearly.... We must be on our guard against those physiological appearances which are proverbially deceptive!"

"It is possible and conceivable that every animal should have been constructed upon a plan of its own, having no resemblance whatever to the plan of any other animal. For any reason we can discover to the contrary, that combination of natural forces which we term life might have resulted from, or been manifested by, a series of infinitely diverse structures; nor would anything in the nature of the case lead us to suspect a community of organization between animals so different in habit and in appearance as a porpoise and a gazelle, an eagle and a crocodile, or a butterfly and a lobster. Had animals been thus independently organized, each working out its life by a mechanism peculiar to itself, such a classification as that now under contemplation would be obviously impossible; a morphological or structural classification plainly implying morphological or structural resemblances in the things classified.

"As a matter of fact, however, no such mutual independence of animal forms exists in nature. On the contrary, the members of the animal kingdom, from the highest to the lowest, are marvelously connected. Every animal has something in common with all its fellows--much with many of them, more with a few, and usually so much with several that it differs but little from them.

"Now, a morphological classification is a statement of these gradations of likeness which are observable in animal structures, and its objects and uses are manifold. In the first place, it strives to throw our knowledge of the facts which underlie, and are the cause of, the similarities discerned into the fewest possible general propositions, subordinated to one another, according to their greater or less degree of generality; and in this way it answers the purpose of a _memoria technica_, without which the mind would be incompetent to grasp and retain the multifarious details of anatomical science."

=Coues on Classification.=--It is obvious that fishes like other animals may be classified in numberless ways, and as a matter of fact by numberless men they have been classified in all sorts of fashions. "Systems," again quoting from Dr. Coues, "have been based on this and that set of characters and erected from this or that preconception in the mind of the systematist.... The mental point of view was that every species of bird (or of fish) was a separate creature, and as much of a fixture in nature's museum as any specimen in a naturalist's cabinet. Crops of classifications have been sown in the fruitful soil of such blind error, but no lasting harvest has been reaped.... The genius of modern taxonomy seems to be so certainly right, to be tending so surely even if slowly in the direction of the desired consummation, that all differences of opinion we hope will soon be settled, and defect of knowledge, not perversity of mind, is the only obstacle in the way of success. The taxonomic goal is not now to find the way in which birds (or other animals) may be most conveniently arranged, but to discover their pedigree, and so construct their family tree. Such a genealogical table, or _phylum_ (~phylon~, tribe, race, stock), as it is called, is rightly considered the only taxonomy worthy the name--the only true or natural classification. In attempting this end, we proceed upon the belief that, as explained above, all birds, like all other animals and plants, are related to each other genetically, as offspring are to parents, and that to discover their generic relations is to bring out their true affinities--in other words, to reconstruct the actual taxonomy of nature. In this view there can be but one 'natural' classification, to the perfecting of which all increase in our knowledge of the structure of birds infallibly and inevitably tends. The classification now in use or coming into use is the result of our best endeavors to accomplish this purpose, and represents what approach we have made to this end. It is one of the great corollaries of that theorem of evolution which most naturalists are satisfied has been demonstrated. It is necessarily a _morphological classification_; that is, one based solely upon considerations of structure or form (~morphê~, form, _morphe_), and for the following reasons: Every offspring tends to take on precisely the form or structure of its parents, as its natural physical heritage; and the principle involved, or the _law of heredity_, would, if nothing interfered, keep the descendants perfectly true to the physical characters of their progenitors; they would 'breed true' and be exactly alike. But counter influences are incessantly operative, in consequence of constantly varying external conditions of environment; the plasticity of organization of all creatures rendering them more or less susceptible of modifications by such means, they become unlike their ancestors in various ways and to different degrees. On a large scale is thus accomplished, by natural selection and other natural agencies, just what man does in a small way in producing and maintaining different breeds of domestic animals. Obviously, amidst such ceaselessly shifting scenes, degrees of likeness or unlikeness of physical structure indicate with the greatest exactitude the nearness or remoteness of organisms in kinship. Morphological characters derived from the examination of structure are therefore the surest guides we can have to the blood relationships we desire to establish; and such relationships are the 'natural affinities' which all classification aims to discover and formulate."

=Species as Twigs of a Genealogical Tree.=--In another essay Dr. Coues has compared species of animals to "the twigs of a tree separated from the parent stem. We name and arrange them arbitrarily in default of a means of reconstructing the whole tree according to nature's ramifications." If one had a tree, all in fragments, pieces of twig and stem, some of them lost, some destroyed, and some not yet separated from the mass not yet picked over, and wished to place each part where he could find it, he would be forced to adopt some system of natural classification. In such a scheme he would lay those parts together which grew from the same branch. If he were compelled to arrange all the fragments in a linear series, he would place together those of one branch, and when these were finished he would begin with another. If all this were a matter of great importance and extending over years or over many lifetimes, with many errors to be made and corrected, a set of names would be adopted--for the main trunk, for the chief branches, the lesser branches, and on down to the twigs and buds.

A task of this sort on a world-wide scale is the problem of systematic zoology. There is reason to believe that all animals and plants sprang from a single stock. There is reasonable certainty that all vertebrate animals are derived from a single origin. These vertebrate animals stand related to each other, like the twigs of a gigantic tree of which the lowermost branches are the aquatic forms to which we give the name of fishes. The fishes are here regarded as composed of six classes or larger lines of descent. Each of these, again, is composed of minor divisions called orders. The different species or ultimate kinds of animals are grouped in genera. A genus is an assemblage of closely related species grouped around a central species as type. The type of a genus is, in common usage, that species with which the name of the genus was first associated. The name of the genus as a noun, often with that of the species which is an adjective in signification if not in form, constitutes the scientific name of the species. Thus _Petromyzon_ is the genus of the common large lamprey, _marinus_ is its species, and the scientific name of the species is _Petromyzon marinus_. _Petromyzon_ means stone-sucker; _marinus_, of the sea, thus distinguishing it from a species called _fluviatilis_, of the river. In like fashion all animals and plants are named in scientific record or taxonomy. Technical names are necessary because vernacular names fail. Half a million kinds of animals are known, while not half a thousand vernacular names exist in any language. And these are always loosely used, half a dozen of them often for the same species, one name often for a dozen species.

In the same way, whenever we undertake an exact description, we must use names especially devised for that purpose. We cannot use the same names for the bones of the head of a fish and those of the head of a man, for a fish has a different series of bones, and this series is different with different fishes.

=Nomenclature.=--A family in zoology is an assemblage of related genera. The name of a family, for convenience, always ends in the patronymic _idæ_, and it is always derived from the leading genus, that is, the one best known or earliest studied. Thus all lampreys constitute the family _Petromyzonidæ_. An order may contain one or more families. An order is a division of a larger group; a family an assemblage of related smaller groups. Intermediate groups are often recognized by the prefixes sub or super. A subgenus is a division of a genus. A subspecies is a geographic race or variation within a species; a super-family a group of allied families. Binomial nomenclature, or the use of the name of genus and species as a scientific name, was introduced into science as a systematic method by Linnæus. In the tenth edition of his Systema Naturæ, published in 1758, this method was first consistently applied to animals. By common consent the scientific naming of animals begins with this year, and no account is taken of names given earlier, as these are, except by accident, never binomial. Those authors who wrote before the adoption of the rule of binomials and those who neglected it are alike "ruled out of court." The idea of genus and species was well understood before Linnæus, but the specific name used was not one word but a descriptive phrase, and this phrase was changed at the whim of the different authors.

=Nomenclature of Trunkfishes.=--Examples of such names are those of the West Indian trunkfish, or cuckold (_Ostracion tricorne_, Linnæus). Lister refers to a specimen in 1686 as "_Piscis triangularis capiti cornutu cui e media cauda cutanea aculeus longus erigitus_." This Artedi alters in 1738 to _Ostracion triangulatus aculeis duobus in capite et unico longiore superne ad caudam_. This is more accurately descriptive and it recognizes the existence of a generic type, _Ostracion_, or trunkfish, to cover all similar fishes. French writers transformed this into various phrases beginning "Coffre triangulaire à trois cornes," or some similar descriptive epithet, and in English or German it was likely to wander still farther from the original. But Linnæus condenses it all in the word _tricornis_, which, although not fully descriptive, is still a name which all future observers can use and recognize.

It is true that common consent fixes the date of the beginning of nomenclature at 1758. But to this there are many exceptions. Some writers date genera from the first recognition of a collective idea under a single name. Others follow even species back through the occasional accidental binomials. Most British writers have chosen the final and completed edition of the Systema Naturæ, the last work of Linnæus, in 1766, in preference to the earlier volume. But all things considered, justice and convenience alike seem best served by the use of the edition of 1758.

=Synonymy and Priority.=--Synonymy is the record of the names applied at different times to the same group or species. With characteristic pungency Dr. Coues defines synonymy as "a burden and a disgrace to science." It has been found that the only way to prevent utter confusion is to use for each genus or species the first name applied to it and no other. The first name, once properly given, is sacred because it is the right name. All other later names whatever their appropriateness are wrong names. In science, of necessity, a name is a name without any necessary signification. For this reason and for the further avoidance of confusion, it remains as it was originally spelled by the author, obvious misprints aside, regardless of all possible errors in classical form or meaning. The names in use are properly written in Latin or in Latinized Greek, the Greek forms being usually preferred as generic names, the Latin adjectives for names of species. Many species are named in honor of individuals, these names being usually given the termination of the Latin genitive, as _Sebastodes gillii, Liparis agassizi_. In recent custom all specific names are written with the small initial; all generic names with the capital.

One class of exceptions must be made to the law of priority. No generic name can be used twice among animals, and no specific name twice in the same genus. Thus the name _Diabasis_ has to be set aside in favor of the next name _Hæmulon_, because _Diabasis_ was earlier used for a genus of beetles. The specific name _Pristipoma humile_ is abandoned, because there was already a _humile_ in the genus _Pristipoma_.

=The Conception of Genus.=--In the system of Linnæus, a genus corresponds roughly to the modern conception of a family. Most of the primitive genera contained a great variety of forms, as well as usually some species belonging to other groups disassociated from their real relationships.

As greater numbers of species have become known the earlier genera have undergone subdivision until in the modern systems almost any structural character not subject to intergradation and capable of exact definition is held to distinguish a genus. As the views of these characters are undergoing constant change, and as different writers look upon them from different points of view, or with different ideas of convenience, we have constant changes in the boundaries of genera. This brings constant changes in the scientific names, although the same specific name should be used whatever the generic name to which it may be attached. We may illustrate these changes and the burden of synonymy as well by a concrete example.

=The Trunkfishes.=--The horned trunkfish, or cuckold, of the West Indies was first recorded by Lister in 1686, in the descriptive phrase above quoted. Artedi, in 1738, recognized that it belonged with other trunkfishes in a group he called _Ostracion_. This, to be strictly classic, he should have written _Ostracium_, but he preferred a partly Greek form to the Latin one. In the Nagg's Head Inn in London, Artedi saw a trunkfish he thought different, having two spines under the tail, while Lister's figure seemed to show one spine above. This Nagg's Head specimen Artedi called "_Ostracion triangulatus duobus aculeis in fronte et totidem in imo ventre subcaudalesque binis_."

Next came Linnæus, 1758, who named Lister's figure and the species it represented, _Ostracion tricornis_, which should in strictness have been _Ostracion tricorne_, as ~ostrakion~, a little box, is a neuter diminutive. The Nagg's Head fish he named _Ostracion quadricornis_. The right name now is _Ostracion tricornis_, because the name _tricornis_ stands first on the page in Linnæus' work, but _Ostracion quadricornis_ has been more often used by subsequent authors because it is more truthful as a descriptive phrase. In 1798, Lacépède changed the name of Lister's fish to _Ostracion listeri_, a needless alteration which could only make confusion.

In 1818, Dr. Samuel Latham Mitchill, receiving a specimen from below New Orleans, thought it different from _tricornis_ and _quadricornis_ and called it _Ostracion sexcornutus_; Dr. Holard, of Paris, in 1857, named a specimen _Ostracion maculatus_, and at about the same time Bleeker named two others from Africa which seem to be the same thing, _Ostracion guineensis_ and _Ostracion gronovii_. Lastly, Poey calls a specimen from Cuba _Acanthostracion polygonius_, thinking it different from all the rest, which it may be, although my own judgment is otherwise. This brings up the question of the generic name. Among trunkfishes there are four-angled and three-angled kinds, and of each form there are species with and without horns and spines. The original _Ostracion_ of Linnæus we may interpret as being _Ostracion cubicus_ of the coasts of Asia, a species similar to the _Ostracion rhinorhynchus_. This species, _cubicus_, we call the type species of the genus, as the Nagg's Head specimen of Artedi was the type specimen of the species _quadricornus_, and the one that was used for Lister's figure the type specimen of _tricornis_.

_Ostracion cubicus_ is a four-angled species, and when the trunkfishes were regarded as a family (_Ostraciidæ_), the three-angled ones were set off as a separate genus. For this two names were offered, both by Swainson in 1839. For _trigonus_, a species without horns before the eyes, he gave the name _Lactophrys_, and for _triqueter_, a species without spines anywhere, the name of _Rhinesomus_. Most recent American authors have placed the three-cornered species which are mostly American in one genus, which must therefore be called _Lactophrys_. Of this name _Rhinesomus_ is a synonym, and our species should stand as _Lactophrys tricornis_. The fact that _Lactophrys_ as a word (from Latin _lætus_, smooth; Greek ~ophrys~, eyebrow; or else from _lactoria_, a milk cow, and ~ophrys~) is either meaningless or incorrectly written makes no difference with the necessity for its use.

In 1862, Bleeker undertook to divide these fishes differently. Placing all the hornless species, whether three-angled or four-angled, in _Ostracion_, he proposed the name _Acanthostracion_ for the species with horns, _tricornis_ being the type. But _Acanthostracion_ has not been usually adopted except as the name of a section under _Lactophrys_. The three-angled American species are usually set apart from the four-angled species of Asia, and our cuckold is called _Lactophrys tricornis_. But it may be with perfect correctness called _Ostracion tricorne_, in the spirit called conservative. Or with the "radical" systematists we may accept the finer definition and again correctly call it _Acanthostracion tricorne_. But to call it _quadricornis_ or _listeri_ or _maculatus_ with any generic name whatever would be to violate the law of priority.

=Trinomial Nomenclature.=--By trinomial nomenclature we mean the use of a second subordinate specific name to designate a geographic subspecies, variety, or other intergrading race. Thus _Salmo clarki virginalis_ indicates the variety of Clark's trout, or the cut-throat trout, found in the lakes and streams of the Great Basin of Utah, as distinguished from the genuine _Salmo clarkii_ of the Columbia. Trinomials are not much used among fishes, as we are not yet able to give many of the local forms correct and adequate definition such as is awarded to similar variations among birds and mammals. Usually varieties in ichthyology count as species or as nothing.

=Meaning of Species.=--Quoting once more from the admirable essay of Dr. Coues on the taxonomy of birds: "The student cannot be too well assured that no such things as species, in the old sense of the word, exist in nature any more than have genera or families an actual existence. Indeed they cannot be, if there is any truth in the principles discussed in our earlier paragraphs. Species are simply ulterior modifications, which once were, if they be not still, inseparably linked together; and their nominal recognition is a pure convention, like that of a genus. More practically hinges upon the way we regard them than turns upon our establishment of higher groups, simply because upon the way we decide in this case depends the scientific labeling of specimens. If we are speaking of a robin, we do not ordinarily concern ourselves with the family or order it belongs to, but we do require a technical name for constant use. That name is compounded of its genus, species, and variety. No infallible rule can be laid down for determining what shall be held to be a species, what a conspecies, subspecies, or variety. It is a matter of tact and experience, like the appreciation of the value of any other group in zoology. There is, however, a convention upon the subject, which the present workers in ornithology in this country find available; at any rate we have no better rule to go by. We treat as "specific" any form, however little different from the next, that we do not know or believe to intergrade with that next one, between which and the next one no intermediate equivocal specimens are forthcoming, and none, consequently, are supposed to exist. This is to imply that differentiation is accomplished, the links are lost and the characters actually become "specific." We treat as "varietal" of each other any forms, however different in their extreme manifestation, which we know to intergrade, having the intermediate specimens before us, or which we believe with any good reason do intergrade. If the links still exist, the differentiation is still incomplete, and the characters are not specific, but only varietal, in the literal sense of these terms."

=Generalization and Specialization.=--A few terms in common use may receive a moment's discussion. A type or group is said to be specialized when it has a relatively large number of peculiarities or when some one peculiarity is carried to an extreme. A sculpin is a specialized fish having many unusual phases of development, as is also a swordfish, which has a highly peculiar structure in the snout. A generalized type is one with fewer peculiarities, as the herring in comparison with the sculpin. In the process of evolution generalized types usually give place to specialized ones. Generalized types are therefore as a rule archaic types. The terms high and low are also relative, a high type being one with varied structure and functions. Low types may be primitively generalized, as the lancelet in comparison with all other fishes, or the herring in comparison with the perch, or they may be due to degradation, a loss of structures which have been elaborately specialized in their ancestry. The sea-snail (_Liparis_), an ally of the sculpin, with scales lost and fins deteriorated is an example of a low type which is specialized as well as degraded.

=High and Low Forms.=--In the earlier history of ichthyology much confusion resulted from the misconception of the terms "high" and "low." Because sharks appeared earlier than bony fishes, it was assumed that they should be lower than any of their subsequent descendants. That the brain and muscular system in sharks was more highly developed than in most bony fishes seemed also certain. Therefore it was thought that the teleost series could not have had a common origin with the series of sharks. It is now understood that evolution means chiefly adaptation. The teleost is adapted to its mode of life, and to that end it is specialized in fin and skeleton rather than in brain and nerves. All degeneration is associated with specialization. The degeneration of the blindfish is a specialization for better adaptation to life in the darkness of caves; the degeneration of the deep-sea fish meets the demands of the depths, the degeneration of the globefish means the sinking of one line of functions in the extension of some other.

Referring to his own work on the fossil fishes in the early forties, Professor Agassiz once said to the writer: "At that time I was on the verge of anticipating the views of Darwin, but it seemed to me that the facts were contrary to the theories of evolution. We had the highest fishes first." This statement leads us to consider what is meant by high and low. Undoubtedly the sharks are higher than the bony fishes in the sense of being nearer to the higher vertebrates. In brain, muscle, teeth, and reproductive structures they are also more highly developed. In all skeletal and cranial characters the sharks stand distinctly lower. But the essential fact, so far as evolution is concerned, is not that the sharks are high or low. They are, in almost all respects, distinctly generalized and primitive. The bony fishes are specialized in various ways through adaptation to the various modes of life they lead. Much of this specialization involves corresponding degeneration of organs whose functions have ceased to be important. As a broad proposition it is not true that "we had our highest fishes first," for in a complete definition of high and low, the specialized perch or bass stands higher. But whether true or not, it does not touch the question of evolution which is throughout a process of adaptation to conditions of life.

Referring to the position of Agassiz and his early friend and disciple, Hugh Miller, Dr. Traquair (1900) uses these words in an address at Bradford, England:

"It cannot but be acknowledged that the paleontology of fishes is not less emphatic in the support of descent than that of any other division of the animal kingdom. But in former days the evidence of fossil ichthyology was by some read otherwise.

"It is now a little over forty years since Hugh Miller died: he who was one of the first collectors of the fossil fishes of the Scottish old red sandstone, and who knew these in some respects better than any other man of his time, not excepting Agassiz himself. Yet his life was spent in a fierce denunciation of the doctrine of evolution, then only in its Lamarckian form, as Darwin had not yet electrified the world with his 'Origin of Species.' Many a time I wonder greatly what Hugh Miller would have thought had he lived a few years longer, so as to have been able to see the remarkable revolution which was wrought by the publication of that book.

"The main argument on which Miller rested was the 'high' state of organization of the ancient fishes of the Paleozoic formations, and this was apparently combined with a confident assumption of the completeness of the geological record. As to the first idea, we know of course that evolution means the passage from the more general to the more special, and that as the general result an onward advance has taken place; yet 'specialization' does not always or necessarily mean 'highness' of organization in the sense in which the term is usually employed. As to the idea of the perfection of the geological record, that of course is absurd.

"We do not and cannot know the oldest fishes, as they would not have had hard parts for preservation, but we may hope to come to know many more old ones, and older ones still than we do at present. My experience on the subject of fossil ichthyology is that it is not likely to become exhausted in our day.

"We are introduced at a period far back in geological history to certain groups of fishes, some of which certainly are high in organization as animals, but yet of generalized type, being fishes and yet having the potentiality of higher forms. But because their ancestors are unknown to us, that it is no evidence that they did not exist, and cannot overthrow the morphological testimony in favor of evolution with which the record actually does furnish us. We may therefore feel very sure that fishes or 'fish-like vertebrates' lived long ages before the oldest forms with which we are acquainted came into existence.

"The modern type of bony fishes, though not so 'high' in many anatomical points as that of the Selachii, Crossopterygii, Dipnoi, Acipenseroidei, and Lepidosteoidei of the Palæozoic and Mesozoic eras, is more specialized in the direction of the fish proper, and, as already indicated, specialization and 'highness' in the ordinary sense of the word are not necessarily coincident. But ideas about these things have undergone a wonderful change since those pre-Darwinian days, and though we shall never be able fully to unravel the problems concerning the descent of animals, we see many things a great deal more clearly now than we did then."

Dr. Gill observes: "Perhaps there are no words in science that have been productive of more mischief and more retarded the progress of biological taxonomy than those words pregnant with confusion, High and Low, and it were to be wished that they might be erased from scientific terminology. They deceive the person to whom they are addressed. They insensibly mislead the one who uses them. Psychological prejudices and fancies are so inextricably associated with these words that the use of them is provocative of such ideas. The words, generalized and specialized, having become almost limited to the expression of the ideas which the scientific biologist wishes to unfold by the others, can with great gain be employed in their stead." ("Families of Fishes," 1872.)

=The Problem of the Highest Fishes.=--As to which fishes should be ranked highest and which lowest, Dr. Gill gives ("Families of Fishes," 1872) the following useful discussion: "While among the mammals there is almost universal concurrence as to the forms entitled to the first as well as the last places, naturalists differ much as to the 'highest' of the ichthyoid vertebrates, but are all of one accord respecting the form to be designated as the 'lowest.' With that admitted lowest form as a starting-point, inquiry may be made respecting the forms which are successively _most nearly related_.

"No dissent has ever been expressed from the proposition that the Leptocardians (_Branchiostoma_) are the lowest of the vertebrates; while they have doubtless deviated much from the representatives of the immediate line of descent of the higher vertebrates, and are probably specialized considerably, in some respects, in comparison with those vertebrates from which they (in common with the higher forms) have descended, they undoubtedly have diverged far less, and furnish a better hint as to the protovertebrates than any other form.

"Equally undisputed it is that most nearly related to the Leptocardians are the Marsipobranchiates (_Lampreys_, etc.), and the tendency has been rather to overlook the fundamental differences between the two, and to approximate them too closely, than the reverse.

"But here unanimity ends, and much difference of opinion has prevailed with respect to the succession in the system of the several subclasses (by whatever name called) of true fishes: (1) Some (e.g., Cuvier, J. Müller, Owen, Lütken, Cope) arranging next to the lowest the Elasmobranchiates, and, as successive forms, the Ganoids and Teleosteans; (2) while others (e.g., Agassiz, Dana, Duméril, Günther) adopt the sequence Leptocardians, Marsipobranchiates, Teleosteans, Ganoids, and Elasmobranchiates. The source of this difference of opinion is evident and results partly from metaphysical or psychological considerations, and partly from those based (in the case of the Ganoids) on real similarities and affinities.

"The evidence in favor of the title of the Elasmobranchiates to the 'highest' rank is based upon (1) the superior development of the brain; (2) the development of the egg, and the ovulation; (3) the possession of a placenta; and (4) the complexity of the organs of generation.

"(1) It has not been definitely stated wherein the superior development of the brain consists, and as it is not evident to the author, the vague claim can only be met by this simple statement; it may be added, however, that the brains comparable in essentials and most similar as a whole to those of the Marsipobranchiates are those of the sharks. In answer to the statement that the sharks exhibit superior intelligence, and thus confirm the indications of cerebral structure, it may be replied that the impression is a subjective one, and the author has not been thus influenced by his own observations of their habits. Psychological manifestations, at any rate, furnish too vague criteria to be available in exact taxonomy.

"(2) If the development of the eggs, their small number, and their investment in cases are arguments in favor of the high rank of the Elasmobranchiates, they are also for the Marsipobranchiates, and thus prove too much or too little for the advocates of the views discussed. The variation in number of progeny among true fishes (e.g., Cyprinodonts, _Embiotocids_) also demonstrates the unreliability of those modifications _per se_.

"(3) The so-called placenta of some Elasmobranchiates may be _analogous_ to that of mammals, but that it is not _homologous_ (i.e., homogenetic) is demonstrable from the fact that all the forms intervening between them and the specialized placental mammals are devoid of a placenta, and by the variation (presence or want) among the Elasmobranchiates themselves.

"(4) The organs of generation in the Elasmobranchiates are certainly more complex than in most other fishes, but as the complexity results from specialization of parts _sui generis_ and different from those of the higher (quadruped) vertebrates, it is not evident what bearing the argument has. If it is claimed simply on the ground of specialization, irrespective of homological agreement with admitted higher forms, then are we equally entitled to claim any specialization of parts as evidence of high rank, or at least we have not been told within what limits we should be confined. The Cetaceans, for example, are excessively specialized mammals, and, on similar grounds, would rank above the other mammals and man; the aye-aye exhibits in its dentition excessive specialization and deviation from the primitive type (as exhibited in its own milk teeth) of the Primates, and should thus also rank above man. It is true that in other respects the higher primates (even including man) may be more specialized, but the specialization is not as obvious as in the cases referred to, and it is not evident how we are to balance _irrelative_ specializations against each other, or even how we shall subordinate such cases. We are thus compelled by the _reductio ad absurdum_ to the confession that irrelative specialization of single organs is untrustworthy, and are fain to return to that better method of testing affinities by the equation of agreement in whole and after the elimination of special teleological modifications.

"The question then recurs, What forms are the most _nearly allied_ to the Marsipobranchiates, and what show the closest approach in _characteristic_ features? And in response thereto the evidence is not undecisive. Wide as is the gap between Marsipobranchiates and fishes, and comparatively limited as is the range of the latter among themselves, the Elasmobranchiates are very appreciably more like, and share more characters in common with them, than any other; so much is this the case that some eminent naturalists (e.g., Pallas, Geoffroy, St. Hilaire, Latreille, Agassiz, formerly Lütken) have combined the two forms in a peculiar group, contradistinguished from the other fishes. The most earnest and extended argument in English, in favor of this combination has been published by Professor Agassiz in his 'Lake Superior,' but that eminent naturalist subsequently arrived at the opposite conclusions already indicated.

"The evidences of the closer affinity of the Elasmobranchiates (than of any other fishes) with the Marsipobranchiates are furnished by (1) the cartilaginous condition of the skeleton; (2) the post-cephalic position of the branchiæ; (3) the development of the branchiæ and their restriction to special chambers; (4) the larger number of the branchiæ; (5) the imperfect development of the skull; (6) the mode of attachment of the teeth; (7) the slight degree of specialization of the rays of the fins; and (8) the rudimentary condition of the shoulder-girdle."

FOOTNOTES:

[147] Key to North American Birds.