io6 



SCIENCE. 



[Vol. XVIII. No. 446 



the separation of the group into several families. The tropi- 

 cal forms composing the family of Blenniidm have from 28 

 to 49 vertebrae, while in the arctic genera the numbers range 

 from 75 to 100. 



The anacanthine fishes in whole or in part seem to have 

 sprung from a blennioid stock. Of these the most specialized 

 group is that of the flounders {Pleuronectidce), already de- 

 scribed. The wide distribution of this family, its members 

 being found on the sandy shores of the zones, renders it es- 

 pecially important in the present discussion. The other 

 anacanthine families are chiefly confined to the cold waters 

 or to the depth of the seas. In the cod family (Gadidce) the 

 number of vertebrse is usually about 50, and in their deep- 

 sea allies, the grenadiers or rat-tails, the numbers range from 

 65 to 80. 



Of the families confined strictly to the fresh waters, the 

 great majority are among the soft-rayed or physostomous 

 fishes, the allies of the salmon, pike, carp, and cat-fish. In 

 all of these the vertebras are numerous. A few fresh- water 

 families have their affinities entirely with the more special- 

 ized forms of the tropical seas. Of these the Centrarchidw 

 (comprising the American fresh-water sun-fish and black 

 bass) have on the average about 30 vertebrae, the pirate perch 

 29, and the perch family, perch and darters, etc., 35 to 45, 

 while the Serranidw or sea bass, the nearest marine relatives 

 of all these, have constantly 24. The marine family of 

 demoiselles {Pomacentridce) have 26 vertebras, while 30 to 

 40 vertebrae usually exist in their fresh-water analogues (or 

 possibly descendants), the Cichlidce, of the rivers of South 

 America and Africa. The sticklebacks, a family of spiny 

 fishes, confined to the rivers and seas of the north, have from 

 31 vertebrae to 41. 



It is apparently true that among the free-swimming, or 

 migratory pelagic fishes, the number of vertebras is greater 

 than among their relatives of local habits. This fact is most 

 evident among the Scombriform fishes, the allies of the mack- 

 erel and tunny. All of these belong properly to the warm 

 seas, and the reduction of the vertebras in certain forms has 

 no evident relation to the temperature, though it seems to 

 be related in some degree to the habits of the species. Per- 

 haps the retention of many segments is connected with that 

 strength and swiftness in the water for which the mackerels 

 are pre-eminent. 



The variations in the number of vertebrae in this group led 

 Dr. Giinther, nearly thirty years ago, to divide it into two 

 families, the Carangidce and Scombrides. The CarangidcB 

 are tropical shore fishes, local or migratory to a slight de- 

 gree. All these have from 24 to 26 vertebrte. In their pelagic 

 relatives, the dolphins, there are from 30 to 33; in the opahs, 

 45; in the braraa, 42; while the great mackerel family, all of 

 whose members are more or less pelagic, have from 31 to 50. 

 Other mackerel-like fishes are the cutlass fishes, which ap- 

 proach the eels in form and in the reduction of the fins. In 

 these the vertebrae are correspondingly numerous, the num- 

 bers ranging from 100 to 160. In apparent contradistinction 

 to this rule, however, the pelagic family of sword-fishes, re- 

 motely allied to the mackerels, and with even greater powers 

 of swimming, has the vertebrae in normal number, the com- 

 mon sword-fish having but 24. 



The eels constitute a peculiar group of uncertain but 

 probably soft-rayed ancestry, in which everything else has 

 been subordinated to muscularity and flexibility of body. 

 The fins, girdles, gill arches, scales, and membrane bones are 

 all imperfectly developed or wanting. The eel is perhaps as 

 far from the primitive stock as the most highly ichthyized 



fishes, but its progress has been of another character. The 

 eel would be regarded in the ordinary sense as a degenerate 

 type, for its bony structure is greatly simplified as compared 

 with its ancestral forms, but in its eel-like qualities it is, how- 

 ever, greatly specialized. All the eels have vertebra in great 

 numbers. As the great majority of the species are tropical, 

 and as the vertebrae in very few of the deep-sea forms have 

 been counted, no conclusions can be drawn as to the relation 

 of their vertebras to the temperature. 



It is evident that the two families most decidedly tropical 

 in their distribution, the morays and the snake-eels, have 

 diverged farthest from the primitive stock. They are most 

 " degenerate," as shown by the reduction of their skeleton. 

 At the same time they are also most decidedly "eel-like," 

 and in some respects, as in coloration, dentition, muscular 

 development, most highly specialized. It is evident that the 

 presence of numerous vertebral joints is essential to the sup- 

 pleness of body which is the eel's chief source of power. So 

 far as known, the niimbers of vertebras in eels range from 

 115 to 160, some of the deep-sea eels having probably higher 

 numbers, if we can draw inferences from their slender or 

 whip-like forms; but this character may be elusive. 



The sharks show likewise a very large number of Tertebrae, 

 130 to 150 in the species in which they have been counted. 

 In these fishes no comparative study of the vertebras has been 

 made. The group is a very ancient one in geological time, 

 and in the comparatively few remaining members of the 

 group, the vertebras, in fact the entire skeleton, is in a very 

 primitive condition. The sharks are free-swimming fishes, 

 and with them as with the eels, flexibility of body is essential 

 to the life they lead. 



In some families the number of rays in the dorsal and 

 anal fins is dependent on the number of vertebrse. It is 

 therefore subject to the same fluctuations. This relation is 

 not strictly proportionate, for often a variable number of 

 rays with their interspinal processes will be interposed be- 

 tween a pair of vertebrae. The myotomes or muscular bands 

 on the sides are usually coincident with the number of ver- 

 tebra. As, however, these and other characters are de- 

 pendent on differences in vertebral segmentation, they bear 

 the same relations to temperature that the vertebrae them- 

 selves sustain. 



From the foregoing examples we may conclude that, other 

 things being equal, the numbers of vertebrae are lowest in 

 the shore-flshes of the tropics, and especially in those of local 

 habits, living about rocks and coral reefs. The cause of 

 this is to be found in the fact that in these localities the in- 

 fluences of natural selection are most active. The production 

 of vertebrae may be regarded as a phase in the process of 

 specialization which has brought about the typical spiny- 

 rayed fish. 



These influences are most active in the warm, clear waters 

 of tropical shores, because these regions offer conditions most 

 favorable to flsh-life, and to the life of the greatest variety 

 of fishes. No fish is excluded from competition. There is the 

 greatest variety of competitors, the greatest variety of fish- 

 food, and the greatest variety of conditions to which adapta- 

 tion is possible. The number of species visiting any single 

 area is vastly greater in the tropics than in cold regions. A 

 single drawing of the net on the shores of Cuba will obtain 

 more different kinds of fish than can be found on the coasts 

 of Maine in a year. Cold, monotony, darkness, isolation, 

 foul water, — all these are characters opposed to the formation 

 of variety in fish-life. The absence of these is a chief feature 

 of life in the tropical waters. 



