"The three-toothed lampreys (Entosphenus tridentatus) of the West Coast climb low falls or rapids by a series of leaps, holding with their mouths to rest, then jumping and striking again and holding, thus leap by leap gaining the entire distance.

"The lampreys here have never been known to show any tendency or ability to climb, probably because there are no rapids or mere low falls in the streams up which they would run. In fact, as the inlet is the only stream entering Cayuga Lake in this region which presents suitable spawning conditions and no obstructions, it can be seen at once that all the lampreys must spawn in this stream and its tributaries.

"In 'running' they move almost entirely at night, and if they do not reach a suitable spawning site by daylight, they will cling to roots or stones during the day and complete their journey the next night. This has been proven by the positive observation of individuals. Of the specimens that run up early in the season, about four-fifths are males. Thus the males do not exactly precede the females, because we have found the latter sex represented in the stream as early in the season as the former, but in the earlier part of the season the number of the males certainly greatly predominates. This proportion of males gradually decreases, until in the middle of the spawning season the sexes are about equally represented, and toward the latter part of the season the females continue to come until they in turn show the greater numbers. Thus it appears very evident in general that the reproductive instinct impels the most of the males to seek the spawning ground before the most of the females do. However, it should be said that neither the males nor the females show all of the entirely sexually mature features when they first run up-stream in the beginning of the season, but later they are perfectly mature and 'ripe' in every regard when they first appear in the stream. When they migrate, they stop at the site that seems to suit their fancy, many stopping near the lake, others pushing on four or five miles farther up-stream. We have noted, however, that later in the season the lower courses become more crowded, showing that the late comers do not attempt to push up-stream as far as those that came earlier. Also it thus follows, from what was just said about late-running females, that in the latter part of the season the lower spawning beds are especially crowded with females. In fact, during the early part of the month of June we have found, not more than half a mile above the lowest spawning bed, as many as five females on a spawning nest with but one male; and in that immediate vicinity many nests indeed were found at that time with two or three females and but one male.

"Having arrived at a shoal which seems to present suitable conditions for a spawning nest, the individual or pair commences at once to move stones with its mouth from the centre to the margin of an area one or two feet in diameter. When many stones are thus placed, especially at the upper edge, and they are cleaned quite free of sediment and algæ, both by being moved and by being fanned with the tail, and when the proper condition of sand is found in the bottom of the basin thus formed, it is ready to be used as a spawning bed or nest. A great many nests are commenced and deserted. This has been left as a mystery in publications on the subject, but we are well convinced that it is because the lampreys do not find the requisites or proper conditions of bottom (rocks, sand, etc., as given below) to supply all their needs and fulfill all conditions for ideal sites. This desertion of half-constructed nests is just what would be expected and anticipated in connection with the explanation of 'Requisite Conditions for Spawning,' given below, because some shallows contain more sand and fewer stones, and others contain many larger stones but no sand, while others contain pebbles lying over either rocks or sand. The lampreys remove some of the material, and if they do not find all the essentials for a spawning nest, the site is deserted and the creatures move on."

Requisite Conditions for Spawning with Lampreys.—"For a spawning site two conditions are immediately essential—proper conditions of water and suitable stream bed or bottom. Of course with these it is essential that no impassable barriers (dam or falls) exist between the lake and the spawning sites to prevent migration at the proper 'running' season. Lampreys will not spawn where there is no sand lying on the bottom between the rocks, as sand is essential in covering the eggs (see remarks on the 'Spawning Process'); neither will they spawn where the bottom is all sand and small gravel, as they cannot take hold of this material with their mouths to construct nests or to hold themselves in the current, and they would not find here pebbles and stones to carry over the nest while spawning, as described elsewhere. It can thus be seen that, as suggested above, the reason they do not spawn in Fall Creek and Cascadilla Creek, between the lake and the falls, is that the beds of these streams are very rocky, being covered only with large stones and no sand. There is no doubt that the lampreys find here suitable conditions of water, but they do not remain to spawn on account of the absence of the proper conditions of stream bed. Again, they do not spawn in the lower course of the inlet for a distance of nearly two miles from the lake, because near the lake the bed of the stream is composed of silt, while for some distance above this (up-stream) there is nothing but sand. Farther up-stream are found pebbles and stones commingled with sand, which combination satisfies the demands of the lampreys for material in constructing nests and covering eggs. The accessibility of these sites, together with their suitable conditions, render the inlet the great and perhaps the only spawning stream of the lake; and, doubtless, all the mature lampreys come here to spawn, excepting a few which spawn in the lower part of Six-mile Creek, a tributary of the inlet.

"As the course of the stream where the beds abound is divided into pools, separated by stony ripples or shallows, the nests must be made at the ends of the pools. Of the spawning beds personally observed during several seasons, nine-tenths of the entire number were formed just above the shallows at the lower ends of the pools, while only a few were placed below them. An advantage in forming the nest above the shoals rather than below it is that in the former place the water runs more swiftly over the lower and middle parts of such a bed than at its upper margin, since the velocity decreases in either direction from the steeper part of the shallows; and any organic material or sediment that would wash over the upper edge of the nest is thus carried on rather than left as a deposit. When formed below the shallows, owing to the decreased velocity at the lower part of the nest compared with that at the upper, the sediment is likely to settle in the hollow of the nest, and, through the process of decay of the organic material, prove disastrous or unfavorable for the developing embryos.

"The necessity of sand in the spawning bed indicates the explanation of why we see so many shallows which have no spawning lampreys upon them, while there are others in the same vicinity that are crowded. There will be no nests formed if there is too little or too much sand, not enough or too many stones, or stones that are all too small or all too large. The stones must vary from the size of an egg to the size of a man's hand, and must be intermingled with sand without mud or rubbish.

"The lampreys choose to make their spawning nests just where the water flows so swiftly that it will carry the sand a short distance, but will not sweep it out of the nest. This condition furnishes not only force to wash the sand over the eggs when laid, but also keeps the adult lampreys supplied with an abundance of fresh water containing the dissolved air needed for their very rapid respiration. Of course in such rapid water the eggs are likely to be carried away down-stream, but Nature provides against this by the fact that they are adhesive, and the mating lampreys stir up the sand with their tails, thus weighing down the freshly laid eggs and holding them in the nest. Hence the necessity of an abundance of sand at the spawning site."

The Spawning Process with Lampreys.—"There is much interest in the study of the spawning process, as it is for the maintenance of the race that the lampreys risk and end their lives; and as they are by far the lowest form of vertebrates found within the United States, a consideration of their actions and apparent evidences of instinct becomes of unusual attraction. Let us consider one of those numerous examples in which the male migrates before the female. When he comes to that portion of the stream where the conditions named above are favorable, he commences to form a nest by moving and clearing stones and making a basin with a sandy bottom about the size of a common wash-bowl. Several nests may be started and deserted before perfect conditions are found for the completion of one. The male may be joined by a female either before or after the nest is completed. There is at once harmony in the family; but if another male should attempt to intrude, either before or after the coming of the female, he is likely to be summarily dealt with and dismissed at once by the first tenant. As soon as the female arrives she too commences to move pebbles and stones with her mouth.

"Sometimes the nest is made large enough to contain several pairs, or often unequal numbers of males and females; or they may be constructed so closely together as to form one continuous ditch across the stream, just above the shallows. Many stones are left at the sides and especially at the upper margin of the nest, and to these both lampreys often cling for a few minutes as though to rest. While the female is thus quiet, the male seizes her with his mouth at the back of her head, clinging as to a fish. He presses his body as tightly as possible against her side, and loops his tail over her near the vent and down against the opposite side of her body so tightly that the sand, accidentally coming between them, often wears the skin entirely off of either or both at the place of closest contact. In most observed instances the male pressed against the right side of the female, although there is no unvarying rule as to position. The pressure of the male thus aids to force the eggs from the body of the female, which flow very easily when ripe. The vents of the two lampreys are thus brought into close proximity, and the conspicuous genital papilla of the male serves to guide the milt directly to the issuing spawn. There appears to be no true intromission, although definite observation of this feature is quite difficult, and, in fact, impossible. During the time of actual pairing, which lasts but a few seconds, both members of the pair exhibit tremendous excitement, shaking their bodies in rapid vibrations and stirring up such a cloud of sand with their tails that their eggs are at once concealed and covered. As the eggs are adhesive and non-buoyant, the sand that is stirred up adheres to them immediately and covers most of them before the school of minnows in waiting just below the nest can dart through the water and regale themselves upon the eggs of these enemies of their race; but woe to the eggs that are not at once concealed. We would suggest that the function of the characteristic anal fin, which is possessed only by the female, and only at this time of year, may be to aid in this vastly important process of stirring up the sand as the eggs are expelled; and the explanation of the absence of such a fin from the ventral side of the tail of the male may be found in the fact that it could not be used for the same purpose at the instant when most needed, since the male is just then using his tail as a clasping organ to give him an essential position in pairing. As soon as they shake together they commence to move stones from one part of the nest to another, to bring more loose sand down over their eggs. They work at this from one to five minutes, then shake again, thus making the intervals between mating from one to five minutes, with a general average of about three and a half minutes.

"Although their work of moving stones does not appear to be systematic in reference to the placing of the pebbles, or as viewed from the standpoint of man, it does not need to be so in order to perfectly fulfill all the purposes of the lampreys. As shown above in the remarks on the spawning habits of the brook lampreys, the important end which they thus accomplish is the loosening and shifting of the sand to cover their eggs; and the more the stones are moved, even in the apparently indiscriminate manner shown, the better is this purpose achieved. Yet, in general, they ultimately accomplish the feat of moving to the lower side of the nest all the stones they have placed or left at the upper margin. At the close of the spawning season when the nest is seen with no large pebbles at its upper margin, but quite a pile of stones below, it can be known that the former occupants completed their spawning process there; but if many small stones are left at the upper edge and at the sides, and a large pile is not formed at the lower edge, it can be known that the nest was forsaken or the lampreys removed before the spawning process was completed. The stones they move are often twice as heavy as themselves, and are sometimes even three or four times as heavy. Since they are not attempting to build a stone wall of heavy material, there is no occasion for their joining forces to remove stones of extraordinary size, and they rarely do so, although once during the past spring (1900) we saw two lake lampreys carrying the same large stone down-stream across their nest. Although this place was occupied by scores of brook lampreys, there were but three pairs of lake lampreys seen here. It is true that one of these creatures often moves the same stone several times, and many even attempt many times to move a stone that has already been found too heavy for it; but sooner or later the rock may become undermined so that the water will aid them, and they have no way of knowing what they can do under such circumstances until they try. Also, the repeated moving of one stone may subserve the same purpose for the lamprey in covering its eggs with sand as would the less frequent removal of many.

"When disturbed on the spawning nest, either of the pair will return to the same nest if its mate is to be found there; but if its mate is in another place, it will go to it, and if its mate is removed or killed, it is likely to go to any part of the stream to another nest. When disturbed, they often start up-stream for a short distance, but soon dart down-stream with a velocity that is almost incredible. They can swim faster than the true fishes, and after they get a start are generally pretty sure to make good their escape, although we have seen them dart so wildly and frantically down-stream that they would shoot clear out on the bank and become an easy victim of the collector. This peculiar kind of circumstance is most likely to happen with those lampreys that are becoming blinded from long exposure to the bright light over the clear running water. If there is a solitary individual on a nest when disturbed, it may not return to that nest, but to any that has been started, or it may stay in the deep pool below the shallows until evening and then move some distance up-stream. When the nest is large and occupied by several individuals, those that are disturbed may return to any other such nest. We have never seen evidence of one female driving another female out of a spawning-nest; and from the great number of nests in which we have found the numbers of the females exceeding those of the males, we would be led to infer that the former live together in greater harmony than do the males.

"Under the subject of the number of eggs laid, we should have said that at one shake the female spawns from twenty to forty. We once caught in fine gauze twenty-eight eggs from a female at one spawning instant. In accordance with the frequency of spawning stated, and the number of eggs contained in the body of one female, the entire length of time given to the spawning process would be from two to four days. This agrees with the observed facts, although the lampreys spend much time in moving stones and thoroughly covering the nests with sand. Even after the work of spawning and moving stones is entirely completed, they remain clinging to rocks in various parts of the stream, until they are weakened by fungus and general debility, when they gradually drift down-stream.

"In forming nests there is a distinct tendency to utilize those sites that are concealed by overhanging bushes, branches, fallen tree-tops, or grass or weeds, probably not only for concealment, but also to avoid the bright sunlight, which sooner or later causes them to go blind, as it does many fishes when they have to live in water without shade. Toward the end of the spawning season, it is very common to see blind lampreys clinging helplessly to any rocks on the bottom, quite unable to again find spawning-beds. However, at such times they are generally spent and merely awaiting the inevitable end.

"As with the brook lamprey, the time of spawning and duration of the nesting period depend upon the temperature of the water, as does also the duration of the period of hatching or development of the embryo. They first run up-stream when the water reaches a temperature of 45° or 48° Fahr., and commence spawning at about 50°. A temperature of 60° finds the spawning process in its height, and at 70° it is fairly completed. It is thus that the rapidity with which the water becomes heated generally determines the length of time the lampreys remain in the stream. This may continue later in the season for those that run later, but usually it is about a month or six weeks from the time the first of this species is seen on a spawning-nest until the last is gone."

What becomes of Lampreys after Spawning?—"There has been much conjecture as to the final end of the lampreys, some writers contending that they die after spawning, others that they return to deep water and recuperate, and yet others compromise these two widely divergent views by saying that some die and others do not. The fact is that the spawning process completely wears out the lampreys, and leaves them in a physical condition from which they could never recover. They become stone-blind; the alimentary canal suffers complete atrophy; their flesh becomes very green from the katabolic products, which find the natural outlet occluded; they lose their rich yellow color and plump, symmetrical appearance; their skin becomes torn, scratched, and worn off in many places, so that they are covered with sores, and they become covered with a parasitic or sarcophytic fungus, which forms a dense mat over almost their entire bodies, and they are so completely debilitated and worn out that recovery is entirely out of the question. What is more, the most careful microscopical examination of ovaries and testes has failed to reveal any evidence of new gonads or reproductive bodies. This is proof that reproduction could not again ensue without a practical rebuilding of the animals, even though they should regain their vitality. A. Mueller, in 1865, showed that all the ova in the lamprey were of the same size, and that after spawning no small reproductive bodies remained to be developed later. This is strong evidence of death after once spawning.

"One author writes that an argument against the theory of their dying after spawning can be found in the fact that so few dead ones have been found by him. However, many can be found dead if the investigator only knows how and where to look for them. We should not anticipate finding them in water that is shallow enough for the bottom to be plainly seen, as there the current is strong enough to move them. It is in the deep, quiet, pools where sediment is depositing that the dead lampreys are dropped by the running water, and there they sink into the soft ooze.

"The absence of great numbers of dead lampreys from visible portions of the stream cannot be regarded as important evidence against the argument that they die soon after spawning once, as the bodies are very soon disintegrated in the water. In the weir that we maintained in 1898, a number of old, worn-out, and fungus-covered lampreys were caught drifting down-stream; some were dead, some alive, and others dying and already insensible, but none were seen going down that appeared to be in condition to possibly regain their strength."

CHAPTER XXIX
THE CLASS ELASMOBRANCHII OR SHARK-LIKE FISHES

The Sharks.—The gap between the lancelets and the lampreys is a very wide one. Assuming the primitive nature of both groups, this gap must represent the period necessary for the evolution of brain, skull, and elaborate sense organs. The interspace between the lampreys and the nearest fish-like forms which follow them in an ascending scale is not less remarkable. Between the lamprey and the shark we have the development of paired fins with their basal attachments of shoulder-girdle and pelvis, the formation of a lower jaw, the relegation of the teeth to the borders of the mouth, the development of separate vertebræ along the line of the notochord, the development of the gill-arches, and of an external covering of enameled points or placoid scales.

These traits of progress separate the Elasmobranchs from all lower vertebrates. For those animals which possess them, the class name of Pisces or fishes has been adopted by numerous authors. If this term is to be retained for technical purposes, it should be applied to the aquatic vertebrates above the lampreys and lancelets. We may, however, regard fish as a popular term only, rather than to restrict the name to members of a class called Pisces. From the bony fishes, on the other hand, the sharks are distinguished by the much less specialization of the skeleton, both as regards form and substance, by the lack of membrane bones, of air-bladder, and of true scales, and by various peculiarities of the skeleton itself. The upper jaw, for example, is formed not of maxillary and premaxillary, but of elements which in the lower fishes would be regarded as belonging to the palatine and pterygoid series. The lower jaw is formed not of several pieces, but of a cartilage called Meckel's cartilage, which in higher fishes precedes the development of a separate dentary bone. These structures are sometimes called primary jaws, as distinguished from secondary jaws or true jaws developed in addition to those bones in the Actinopteri or typical fishes. In the sharks the shoulder-girdle is attached, not to the skull, but to a vertebra at some distance behind it, leaving a distinct neck, such as is possessed or retained by the vertebrate higher than fishes. The shoulder-girdle itself is a continuous arch of cartilage, joining its fellow at the breast of the fish. Other peculiar traits will be mentioned later.

Characters of Elasmobranchs.—The essential character of the Elasmobranchs as a whole are these: The skeleton is cartilaginous, the skull without sutures, and the notochord more or less fully replaced or inclosed by vertebral segments. The jaws are peculiar in structure, as are also the teeth, which are usually highly specialized and found on the jaws only. There are no membrane bones; the shoulder-girdle is well developed, each half of one piece of cartilage, and the ventral fins, with the pelvic-girdle, are always present, always many-rayed, and abdominal in position. The skin is covered with placoid scales, or shagreen, or with bony bucklers, or else it is naked. It is never provided with imbricated scales. The tail is diphycercal, heterocercal, or else it degenerates into a whip-like organ, a form which has been called leptocercal. The gill-arches are 5, 6, or 7 in number, with often an accessory gill-slit or spiracle. The ventral fins in the males (except perhaps in certain primitive forms) are provided with elaborate cartilaginous appendages or claspers. The brain is elongate, its parts well separated, the optic nerves interlacing. The heart has a contractile arterial cone containing several rows of valves; the intestine has a spiral valve; the eggs are large, hatched within the body, or else deposited in a leathery case.

Classification of Elasmobranchs.—The group of sharks and their allies, rays, and Chimæras, is usually known collectively as Elasmobranchii (ἐλάσμος, blade or plate; βράγχος, gill). Other names applied to all or a part of this group are these: Selachii (σελαχός, a cartilage, the name also used by the Greeks for the gristle-fishes or sharks); Plagiostomi (πλαγιός, oblique; στόμα, mouth); Chondropterygii (χόνδρος, cartilage; πτερύξ, fin); and Antacea (ἀντακαῖος, sturgeon). They represent the most primitive known type of jaw-bearing vertebrates, or Gnathostomi (γνάθος, jaw; στόμα, mouth), the Chordates without jaws being sometimes called collectively Agnatha (ἀ-γνάθος, without jaws). These higher types of fishes have been also called collectively Lyrifera, the form of the two shoulder-girdles taken together being compared to that of a lyre. Through shark-like forms all the higher vertebrates must probably trace their descent. Sharks' teeth and fin-spines are found in all rocks from the Upper Silurian deposits to the present time, and while the majority of the genera are now extinct, the class has had a vigorous representation in all the seas, later Palæozoic, Mesozoic, and Cenozoic, as well as in recent times.

Most of the Elasmobranchs are large, coarse-fleshed, active animals feeding on fishes, hunting down their prey through superior strength and activity. But to this there are many exceptions, and the highly specialized modern shark of the type of the mackerel-shark or man-eater is by no means a fair type of the whole great class, some of the earliest types being diminutive, feeble, and toothless.

Subclasses of Elasmobranchs.—With the very earliest recognizable remains it is clear that the Elasmobranchs are already divided into two great divisions, the sharks and the Chimæras. These groups we may call subclasses, the Selachii and the Holocephali, or Chismopnea.

The Selachii, or sharks and rays, have the skull hyostylic, that is, with the quadrate bone grown fast to the palate which forms the upper jaw, the hyomandibular, acting as suspensorium to the lower jaw, being articulated directly to it.

The palato-quadrate apparatus, the front of which forms the upper jaw in the shark, is not fused to the cranium, although it is sometimes articulated with it. There are as many external gill-slits as there are gill-arches (5, 6, or 7), and the gills are adnate to the flesh of their own arches, without free tips. The cerebral hemispheres are grown together. The teeth are separated and usually strongly specialized, being primitively modified from the prickles or other defences of the skin. There is no frontal holder or bony hook on the forehead of the male.

The subclass Holocephali, or Chimæras, differ from the sharks in all this series of characters, and its separation as a distinct group goes back to the Devonian or even farther, the earliest known sharks having little more in common with Chimæras than the modern forms have.

The Selachii.—There have been many efforts to divide the sharks and rays into natural orders. Most writers have contented themselves with placing the sharks in one order (Squali or Galei or Pleurotremi) having the gill-openings on the side, and the rays in another (Rajæ, Batoidei, Hypotrema) having the gill-openings underneath. Of far more importance than this superficial character of adaptation are the distinctions drawn from the skeleton. Dr. Gill has used the attachment of the palato-quadrate apparatus as the basis of a classification. The Opistharthri (Hexanchidæ) have this structure articulated with the postorbital part of the skull. In the Prosarthri (Heterodontidæ) it is articulated with the preorbital part of the skull, while in the other sharks (Anarthri) it is not articulated at all. But these characters do not appear to be always important. Chlamydoselachus, for example, differs in this regard from Heptranchias, which in other respects it closely resembles. Yet, in general, the groups thus characterized are undoubtedly natural ones.

Hasse's Classification of Elasmobranchs.—In 1882, Professor Carl Hasse proposed to subdivide the sharks on the basis of the structure of the individual vertebræ. In the lowest group, a hypothetical order of Polyospondyli, possibly represented by the fossil spines called Onchus, an undivided notochord, perhaps swollen at regular intervals, is assumed to have represented the vertebral column. In the Diplospondyli (Hexanchidæ) the imperfectly segmented vertebræ are joined in pairs, each pair having two neural arches. In the Asterospondyli or ordinary sharks each vertebra has its calcareous lamella radiating star-like from the central axis. In the Cyclospondyli (Squalidæ, etc.) the calcareous part forms a single ring about the axis, and in the Tectospondyli (Squatina, rays, etc.) it forms several rings. These groups again are natural and correspond fairly with those based on other characters. At the same time there is no far-reaching difference between Cyclospondyli and Tectospondyli, and the last-named section includes both sharks and rays.

Nothing is known of the Polyospondyli, and they may never have existed at all. The Diplospondyli do not differ very widely from the earlier Asterospondyli (Cestraciontes) which, as a matter of fact, have preceded the Diplospondyli in point of time, if we can trust our present knowledge of the geological record.

Other Classifications of Elasmobranchs.—Characters more fundamental may be drawn from the structure of the pectoral fin. In this regard four distinct types appear. In Acanthoessus this fin consists of a stout, stiff spine, with a rayless membrane attached behind it. In Cladoselache the fin is low, with a very long base, like a fold of skin (ptychopterygium), and composed of feeble rays. In Pleuracanthus it is a jointed axis of many segments, with a fringe of slender fin-rays, corresponding in structure to all appearance to the pectoral fin of Dipnoans and Crossopterygians, the type called by Gegenbaur archipterygium on the hypothesis that it represents the primitive vertebrate limb.

In most sharks the fin has a fan-shape, with three of the basal segments larger than the others. Of these the mesopterygium is the central one, with the propterygium before it and the metapterygium behind. In the living sharks of the family of Heterodontidæ, this form of fin occurs and the teeth of the same general type constitute the earliest remains distinctly referable to sharks in the Devonian rocks.

Primitive Sharks.—Admitting that these four types of pectoral fin should constitute separate orders, we have next to consider which form is the most primitive and what is the line of descent. In this matter we have, in the phrase of Hæckel, only the "three ancestral documents, Palæontology, Morphology, and Ontogeny."

Unfortunately the evidence of these documents is incomplete and conflicting. So far as Palæontology is concerned, the fin of Cladoselache, with that of Acanthoessus, which may be derived from it, appears earliest, but the modern type of pectoral fin with the three basal segments is assumed to have accompanied the teeth of Psammodonts and Cochliodonts, while the fin of the Chimæra must have been developed in the Devonian. The jointed fin of Cladodus and Pleuracanthus may be a modification or degradation of the ordinary type of shark-fin.

Assuming, however, that the geological record is not perfect and that the fin of Cladoselache is not clearly shown to be primitive, we have next to consider the evidence drawn from morphology.

Those who with Balfour and others (see page 69) accept the theory that the paired fins are derived from a vertebral fold, will regard with Dean the fin of Cladoselache as coming nearest the theoretical primitive condition.

The pectoral fin in Acanthoessus Dean regards as a specialized derivative from a fin like that of Cladoselache, the fin-rays being gathered together at the front and joined together to form the thick spine characteristic of Acanthoessus. This view of the morphology of the fin of Acanthoessus is not accepted by Woodward, and several different suggestions have been recorded.

If with Gegenbaur we regard the paired fins as derived from the septa between the gill-slits, or with Kerr regard them as modified external gills, the whole theoretical relation of the parts is changed. The archipterygium of Pleuracanthus would be the nearest approach to the primitive pectoral limb, and from this group and its allies all the other sharks are descended. This central jointed axis of Pleuracanthus is regarded by Traquair as the equivalent of the metapterygium in ordinary sharks. (See Figs. 44, 45, 46.)

According to Traquair: "The median stern [of the archipterygium], simplified, shortened up and losing all its radials on the postaxial side, except in sometimes a few near the tip, becomes the metapterygium, while the mesopterygium and propterygium are formed by the fusion into two pieces of the basal joints of a number of preaxial radials, which have reached and become attached to the shoulder-girdle in front of the metapterygium."

According to Dr. Traquair, the pectoral fin in Cladodus neilsoni, a shark from the Coal Measures of Scotland, is "apparently a veritable uniserial archipterygium midway between the truly biserial one of Pleuracanthus and the pectoral fin of ordinary sharks." Other authors look on these matters differently, and Dr. Traquair admits that an opposite view is almost equally probable. Cope and Dean would derive the tribasal pectoral of ordinary sharks directly from the ptychopterygium or fan-like fold of Cladoselache, while Fritsch and Woodward would look upon it as derived in turn from the Ceratodus-like fin of Pleuracanthus, itself derived from the ptychopterygium or remains of a lateral fin-fold.

If the Dipnoans are descended from the Crossopterygians, as Dollo has tried to show, the archipterygium of Pleuracanthus has had a different origin from the similar-appearing limb of the Dipnoans, Dipterus and Ceratodus.

In such case the archipterygium would not be the primitive pectoral limb, but a structure which may have been independently evolved in two different groups.

In the view of Gegenbaur, the Crossopterygians and Dipnoans with all the higher vertebrates and the bony fishes would arise from the same primitive stock, ancestors, or allies of the Ichthyotomi, which group would also furnish the ancestors of the Chimæras. In support of this view, the primitive protocercal or diphycercal tail of Pleuracanthus may be brought in evidence as against the apparently more specialized heterocercal tail of Cladoselache. But this is not conclusive, as the diphycercal tail may arise separately in different groups through degeneration, as Dollo and Boulenger have shown.

The matter is one mainly of morphological interpretation, and no final answer can be given. On page 68 a summary of the various arguments may be found. Little light is given by embryology. The evidence of Palæontology, so far as it goes, certainly favors the view of Balfour. Omitting detached fin-spines and fragments of uncertain character, the earliest identifiable remains of sharks belong to the lower Devonian. These are allies of Acanthoessus. Cladoselache comes next in the Upper Devonian. Pleuracanthus appears with the teeth and spines supposed to belong to Cestraciont sharks, in the Carboniferous Age. The primitive-looking Notidani do not appear before the Triassic. For this reason the decision as to which is the most primitive type of shark must therefore rest unsettled for the present and perhaps for a long time to come.

The weight of authority at present seems to favor the view of Balfour, Wiedersheim, Boulenger, and Dean, that the pectoral limb has arisen from a lateral fold of skin. But weight of authority is not sufficient when evidence is confessedly lacking.

For our purpose, without taking sides in this controversy, we may follow Dean in allowing Cladoselache to stand as the most primitive of known sharks, thus arranging the Elasmobranchs and rays, recent and fossil, in six orders of unequal value—Pleuropterygii, Acanthodei, Ichthyotomi, Notidani, Asterospondyli, and Tectospondyli. Of these orders the first and second are closely related, as are also the fourth and fifth, the sixth being not far remote. The true sharks form the culmination of one series, the rays of another, while from the Ichthyotomi the Crossopterygians and their descendants may be descended. But this again is very hypothetical, or perhaps impossible; while, on the other hand, the relation of the Chimæras to the sharks is still far from clearly understood.

Order Pleuropterygii.—The order of Pleuropterygii of Dean (πλεύρον, side; πτερύξ, fin), called by Parker and Haswell Cladoselachea, consists of sharks in which the pectoral and ventral fins have each a very wide horizontal base (ptychopterygium), without jointed axis and without spine. There are no spines in any of the fins. The dorsal fin is low, and there were probably two of them. The notochord is persistent, without intercalary cartilage, such as appear in the higher sharks. The caudal fin is short, broad, and strongly heterocercal. Apparently the ventral fin is without claspers. The gill-openings were probably covered by a dermal fold. The teeth are weak, being modified denticles from the asperities of the skin. The lateral line is represented by an open groove. The family of Cladoselachidæ consists of a single genus Cladoselache from the Cleveland shale or Middle Devonian of Ohio. Cladoselache fyleri is the best-known species, reaching a length of about two feet. Dean regards this as the most primitive of the sharks, and the position of the pectorals and ventrals certainly lend weight to Balfour's theory that they were originally derived from a lateral fold of skin. I am recently informed by Dr. Dean that he has considerable evidence that in Cladoselache the anus was subterminal. If this statement is verified, it would go far to establish the primitive character of Cladoselache.

Order Acanthodei.—Near the Pleuropterygii, although much more highly developed, we may note the strange group of Acanthodei (ἀκανθώδης, spinous). These armed fishes were once placed among the Crossopterygians, but there seems no doubt that Woodward is right in regarding them as a highly specialized aberrant offshoot of the primitive sharks. In this group the paired fins consist each of a single stout spine, nearly or quite destitute of other rays. A similar spine is placed in front of the dorsal fin and one in front of the anal. According to Dean these spines are each produced by the growing together of all the fin-rays normally belonging to the fin, a view of their morphology not universally accepted.

The dermal covering is highly specialized, the shagreen denticles being much enlarged and thickened, often set in little squares suggesting a checker-board. The skull is covered with small bony plates and membrane bones form a sort of ring about the eye. The teeth are few, large, and "degenerate in their fibrous structure." Some of the species have certainly no teeth at all. The tail is always heterocercal, or bent upward at tip as in the Cladoselache, not diphycercal, tapering and horizontal as in the Ichthyotomi.

The lower Acanthodeans, according to Woodward, "are the only vertebrates in which there are any structures in the adult apart from the two pairs of fins which may be plausibly interpreted as remnants of once continuous lateral folds. In Climatius, one of the most primitive genera (see Fig. 305), there exists, according to Woodward, and as first noticed by Cope, between the pectoral and pelvic (or ventral) fins a close and regular series of paired spines, in every respect identical with those supporting the appendages that presumably correspond to the two pairs of fins in the higher genera. They may even have supported fin membranes, though specimens sufficiently well preserved to determine this point have not yet been discovered. However, it is evident that dermal calcifications attained a greater development in the Acanthodei than in any of the more typical Elasmobranchs, and we may look for much additional information on the subject when the great fishes to which the undetermined Ichthyodorulites pertained became known." (See Fig. 305.)

The Acanthodei constitute three families. In the Acanthoessidæ there is but one short dorsal fin opposite the anal, and clavicular bones are absent. The gill-openings being provided with "frills" or collar-like margins, perhaps resembled those of the living genus Chlamydoselachus, the frilled shark. The pectoral spine is very strong, and about the eye is a ring of four plates. The body is elongate, tapering, and compressed. Acanthoessus of Agassiz, the name later changed by its author to Acanthodes, is the principal genus, found in the Devonian and Carboniferous.

The species of Acanthoessus are all small fishes rarely more than a foot long, with very small teeth or none, and with the skin well armed with a coat-of-mail. Acanthoessus bronni is the one longest known. In the earliest species known, from the Devonian, the ventral fins are almost as large as the pectorals and nearly midway between pectorals and anal. In the later species the pectoral fins become gradually larger and the ventrals move forward. In the Permian species the pectorals are enormous.

Traquairia pygmæa, from the Permian of Bohemia, is a diminutive sharklet three or four inches long with large scales, slender spines, and apparently no ventral fins.

In the genus Cheiracanthus the dorsal fin is placed before the anal. In Acanthodopsis the teeth are few, large, and triangular, and the fin-spines relatively large.

The Ischnacanthidæ have no clavicles, and two dorsal fins. Ischnacanthus gracilis of the Devonian has a few large conical teeth with small cusps between them.

The Diplacanthidæ, with two dorsal fins, possess bones interpreted as clavicles. The teeth are minute or absent. In Diplacanthus striatus and Diplacanthus longispinus of the Lower Devonian stout spines are attached to the shoulder-girdle between the pectoral spines below.