662 PISCES. 



has aisappeared from the ocean subsequently to it^ nrst inriTiduction ; imr tr there any that seems to hare uiider- 

 \^onQ ;iny rtiminution ; so that the Ctenoid and Cyclnid Ordurs may be said t'.i be presenting their highest develop- 

 ment at the present time. 



When we .survey the Geologieal distribution of tlie other two f.)rders, liowever, we see a most extraordinary 

 contrast. AUhou^;h they now form so small a part of the inhabitants of our seas, we loulc back to a time when 

 tbey were the sole Vei'tebrated tenants of the globe; and we see that in the period anterior to that of the 

 X'redumiiiance of the great extinct aquatic Reptiles, whorse remains abound in the Lias and Oolite formations, 

 fur-tain ti-ibes of each Order had attained a very high degree of importance. Of several families, moreover, i\'liicli 

 i.xistL'1 in the earlier periods of the history of the globe, some of them even having been the most numerous nirl 

 important tribes in the whole class as then existing, not a single representative now remains. Of these, the mftst 

 remarkable among the Ganoid fishes is the family of X^'pWoMa ; which was characterized by the possession nf 

 numerous rows of brush-like teeth, and by the covering of flat rhomboidal scaler arranged parallel with the body. 

 Remains of this family present themselves in nearly the oldest fossiliferous strata; it first began to abound, how- 

 ever, in the Carboniferous period; attained its fullest development in the period of the Triassic formation; 

 sli;j:litly diminishL'iJ daring Lhe Jurassie period ; u'nU'rwent a still greater diminution during the deposition of the 

 Chalk; and disapiiearc-d c'..mph/t.;ly in tliL- Tertiary u|MX-h. S^'arcely luss remarkable is the history of the family 

 of SavroitJ fi>h, so named from the nuuieruas points of reseniblance to Saurian reptiles, which occur in tlirir 

 internal structure. This family seums to have commenced somewhat later than the preceding, but to liavo 

 attained its fullest development at about the same part of the series. Its existence, however, ha"; been continuL'd 

 downwards to the present time ; though it is now representL-d by only two genera, both of them restricted to fresh 

 nnt.r, — viz. the pDl.vpterus, an inhabitant of the rivers of "Western Africa, and the Lepidosteus, a tenant of the 

 rivL'i-s and lakes uf N'ui'tli America, 



Ii] the I'LvcoiD i.ifdcr, thr family of CcsCr'fC(0>(/5 coi-resj.onds very closely in its history with tlie Sauroid and 

 Lepiddid Ganoideans. This family bears a general resemblance to the Sharks in the form of the body aud in 

 internal structure ; but the teeth, instead of being sharp and lancet-shaped, are flat and pavement-like, adapti.'d 

 for crushing instead of for cutting. Remains of Cestraciont fish are among the earliest that present themsulvus 

 in the Palieozoic rucks ; the family increased in importance through the Carboniferous series, and attained its 

 greatest development in the Triassic ; after which it progressively diminished, and is now represented by only a 

 single species, the Cestraeion Philippi, or Port Jackson Shark. 



The families of Ganoid and Placoid fishes, which are jiou' most numerous, may, for the most part, be traced 

 backwards to the remoter epochs. Thus the Sturgeons and the Bai/s have existtd, nearly in the same proportion 

 as at present, from the beginning of the Lias formation ; th« C'him(croid fish date from the cummencenient of the 

 Secondary period; whilst the Squaloids, or true Sharks, make their first appearance in the seas of the Carboni- 

 ferous epoch, and have been gradually increasing in importance down to the present time. It is very instructive 

 to compare the present predominance of these sharp-toothed Sharks, with the former high development of the 

 Cestracionts or blunt-toothed Sharks ; and to note how closely the gradual increase of the one tribe corresponds 

 with the decrease of the other. When we view these facts in connection with the general condition of the class 

 at each epoch, we find the explanation of it perfectly easy; for the period of highest development of the Cestra- 

 ciont family was that at which nearly all other existing fish were of the Ganoid order; that is, were covered 

 with an armature of bony or enamelled scales or plates, quite impenetrable to any simple mtting instrument, and 

 requiring powerful cvuilnug teeth to make any impression upon them ; \\hilst, on the other hand, it is only since 

 the introduction of the Ctenoid and Cycloid fish, whose thin horny scales present no such impediment, that we 

 find the family of sharp-toothed Shaik-., tu wliiuli thi.y furnish appropriate food, rising into impurtanee. 



Omitting from the Ganoid Order tlif SiJuroid iamily, which more properly belongs to the Cycloids, and omitting 

 from the I'lacoid Order tVie Cyelostome fish, which cannotbe appropriately ranged under any ..me of t!ie pj-ini;irY 

 divisi^ins of professor Agassiz, it may be stated as a general fact, that all the existing Ganoid and Plac<.ud fi-hes 

 are representatives of families, whose first appearance dates back at least as far back as the commeoeement uf 

 the Tertiary epoch, most of them being of much older date. The principal development of these Orders shows 

 itself in the Palaeozoic and Secondary periods, th;it is, during the formation of all rocks older than the Chalk ; 

 and the class being then entirely destiture of Ctenoid and Cycloid fislies, must have possessed, as a whole, a very 

 different aspect from that which it now presents. 



Another general fact of nmch scientific interest is brought into view by the study of the fishes of the older 

 Geological formations. In all the Ctenoid and Cycloid fishes, theeamlalfin is equally expanded above and below. 

 and commences from the end of the vertebral column, -which does not pass into it. This form uf tail is called by 

 Professor Agassiz, the Iu>mocercal tail. Rut in the Sharks, Sturgeons, and Lepidosteus of the present time, we find 

 the caudal fin composed of two unequal branches, of which the upper one is supported upon a prolongation of 

 the vertebral column, whilst the lower and shorter is given oft" from its under side. This form of tail is called 

 the hcterocercal. All the earlier Placoid and Ganoid fish seem to have possessed this latter form of tiil;the 

 homocercal confirmation not manifesting itself until after the commencement of the Secondary ]a:-riiid. Rutli 

 varieties present themselves in the Sauroid and Lepidoid fish ; the heterocercal in the older, and the hmnoeereal 

 in the more recent. It is a curious fact that aii fishes have the heterocercal character of tail at an e.-irly periml 

 of their embryonic development ; so that this may be considered as the more tjrejterai form, wliieli gives ]ilaee, in 

 certain cases, to one more specially adaijtcd to the conditions of their existence. 



Although the classification of Professor Agassiz has thus been of the greatest service to the Geologist, and was 

 a great advance u|jou that of Cuvier as regajds its adafifation to the reception of the extinct forms of the class 

 it lias the faults to which all classifications that are based on s'ukjI.c ch,araeters are liable ; dissimilar tribes beiii''- 



