SKELETON OF COMMON PERCH. 97 



opinion of the late Professor Balfour has been secondarily acquired. The caudal 

 fin is first differentiated from the continuous fold : then the posterior dorsal when 

 present, unless the anterior dorsal is of a peculiar type, as in Lophius, when it 

 appears before the posterior. The anal fin appears before the pelvic fins, unless 

 the latter are of a peculiar type and adapted to special uses, as in the young of some 

 Gadoidei. 



The young Elasmobranch, Lepidosteus, and Teleostean, have at first a long 

 pointed tail, to the tip of which the notochord extends and the lateral line as well 

 whenever this primitive condition is retained. A caudal fin is next developed on 

 the ventral aspect and at some little distance from the extremity of this pointed 

 prolongation as an enlargement of the continuous fold. The prolongation bends 

 upwards towards the dorsal aspect at the same time. The growing caudal fin, as 

 remarked by Professor A. Agassiz, has much the appearance of a second or pos- 

 terior anal. It is supported solely by enlarged haemal arches, beyond which 

 appear at a later period the fin-rays. Such a condition persists and forms the 

 heterocercal caudal fin of all Elasmobranchii, living chondrostean Ganoids, and 

 many extinct Ganoids. But in existing bony Ganoids and the Teleostei the pointed 

 prolongation atrophies until the caudal fin becomes terminal. The upward dorsal 

 inclination is preserved in the urostyle. The part of the caudal fin formed on the 

 neural side of the urostyle is always inconsiderable. This outwardly symmetrical 

 caudal fin, really asymmetrical and, anatomically speaking, heterocercal, is termed 

 homocercal. But in a few fish, such as the Dipnoi and the Teleostei above- 

 mentioned, the backbone retains its straight course and divides the caudal fin into 

 two equal portions, dorsal and ventral. Such fins are known as diphycercal. In 

 the Eel and some other eel-like Teleosteans, rudimentary haemal arches exist 

 and point to the existence at some distant period of a caudal lobe now aborted : 

 and in them the diphycercal tail is secondarily acquired. In Holocephali the long 

 whip-like tail has the groove of the lateral line continued to its apex, and a small 

 ventral lobe represents the large caudal lobe of Elasmobranchii. The late Pro- 

 fessor Balfour believed that he had found traces of caudal haemal arches in Cera- 

 todus, which would indicate a lost caudal lobe in that Dipnoan. It is perhaps 

 doubtful whether a primary diphycercal tail exists among living Pisces. 



The number of haemal arches present in the caudal lobe varies much, e.g. the 

 Cod and Stickleback have only two. The ' accessory ' rays are in some instances 

 of large size, e.g. in the Cod, and the caudal fin has then a peculiar rounded 

 appearance. 



The series of bones known as supra-clavicle, clavicle, inter-clavicle, and post- 

 clavicle in Teleostei and Ganoidei appear one and all to be derived from the integu- 

 ment and mucous membrane of the branchial cavity. In the Teleostei they are dermal 

 in position; in Acipenser on the contrary their outer surface resembles in appearance 

 the bony plates of the integument, and the bone has also the same structure. Re- 

 cent researches appear to have established the fact that the clavicle of higher forms 

 is a process of the cartilaginous shoulder-girdle (see Gotte and Hoffmann's papers 

 referred to under Pigeon, p. 67). Such a process exists in the Sturgeon, and is 

 named prae-coracoid in Professor W. K. Parker's Monograph on the shoulder- 

 girdle. 'Swirski's researches on the shoulder-girdle in the Pike appear to establish 

 the fact that the true coracoid aborts almost completely in that Teleostean, and 



H 



