518 REroPtT or co^oiissioner of fish and FisnEHiES. [04] 



arranging it.^elf in parallel bands from which the cartilaginous matrix 

 of the rays is developed. The rudiments of the bony sheaths of the 

 rays of the caudal fin in sections of that portion of larval salmon are 

 found to lie almost in contact with the skin and to be crescentic in sec- 

 tion. These are evidently the lateral pieces which develop into the 

 bony segmented sheaths of the rays of the adult, which are of much the 

 same form except that they more completely ensheath the cartilaginous 

 matrix of the rays than in the larvae. At the base of the caudal fin of 

 the salmon embryo the sheaths of the fin-rays lie deeper than in its dis- 

 tal portion, and a stratum of tissue is interposed between the skin and 

 the sheaths of the rays, which is afterwards apparently developed into 

 the flexor muscles of this fin. The cartilages, which afterwards ossify 

 and become the hypural bones, are mesial in position and a considerable 

 thickness of tissue is interposed between them and the rudiments of 

 the rays lying on either side. Such, in general terms, appears to be 

 the process of caudal and median fin development in osseous fishes. In 

 some forms there is a tendency manifested to develop more or less 

 mesoblast in the median fin-fold, and vascular loops also appear in its 

 mesial substance at an early stage of development, as may be seen in 

 the embryos of Apeltes and TyloRuriis. In other forms, again, the me- 

 dian fin-fold retains its thin, transparent, dermal, non-vascular character 

 for a long time after hatching ; this is noteworthy in Alosa^ Pomolobus, 

 CyMum, Parcphip'pus, and IduSj while in ylj9f//e9 vascular loops are pres- 

 ent in it by the time of hatching. In Siphostoma and Gamhusia the fin- 

 folds of the unpaired fins j^row out as local dermal folds into which mes- 

 oblastic tissue is almost immediately insinuated to develoi) the rays. 



It will be evident, from what has preceded, that the theory of the 

 origin of the unpaired fins from continuous folds does not hold in Tele- 

 osts; that there are exceptions to it where we should least ex])ect to 

 find them; in fine, that the form of the caudal fin is sometimes outlined 

 before the rays appear or the reverse. It is worth while, however, to 

 l^oint out that in those forms with two or three dorsals {Gadus), or 

 where there is a series of dorsal and ventral finlets in the adult {Cybiiim), 

 the continuous larval fin is most apt to be developed, while in those 

 forms where there is a reduction (degeneration) of the fin system, as in 

 the Lophobranchs and Ganihufiia (with but one dorsal), the continuous 

 median fin-fold is not always developed. The last qualification does 

 not hold, however, in all cases, for in the larva of some Cyi)rinoids, 

 Idns and Caras.sius (one dorsal in the adult), I find the continuous lin- 

 fold developed to the same extent as in Apcltcs ; and amongst the Clu- 

 l)eoids, which have but one short dorsal in the adult, it is surprising- to 

 find the natatory fold extensively developed. The query arises, why 

 should Gamhusia form su(!h an anomalous excei)tion? We can under- 

 stand the cause of the peculiar develoi)mentof the median fins of Lopho- 

 branchs as resulting from their extreme specialization, but in the first 

 case the exi>lanation is not so clear. 



