1046 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [66] 



divergent basal bars, su])ported by tbe antorbital cartilages, partly on 

 the sides and ends of the cranial rostriim. By what process these bars 

 came to assume such a relation to the chondro-cranium of Torpedo is 

 not clear, unless we assume, as supposed by Gegenbaur, that a part of 

 the anterior end of the propterygium be(;ame detached from the pec- 

 toral and was secondarily affixed to the antorbital process of the skull. 

 This is an explanation, but is it the one which will be finally accepted? 

 This may be doubted, for even supposing that this view is the true one, 

 it is rather singular that the cartilaginous rostrum in Hypnos should 

 divide into three lobes, each supporting rays, while on either side of it 

 accessory rays should be found. What seems to be the difficulty is this: 

 Why must it be supposed that the rostral rays are not directly devel- 

 oped in the connection in which they are found'? An answer to this 

 question can scarcely be expected until the development of the cephalic 

 tin of Torpedo or one of its alhes has been actually traced. 



The peculiarities of development of greatly si^ecialized forms evidently 

 do not furnish safe criteria upon which to base general conclusions in 

 respect to the development of all affiliated forms, as the i)receding cases 

 would seem to indicate. 4-^ another instance of this may be cited the 

 fact that the single vertebral centra of Lopliohrancliii support a num- 

 ber of neural arches, which in certain regions are certainly first laid 

 down in cartilage. The usual rule is a single neural arch to a single 

 centrum, though in the tailof Amm this rule is violated in another way, 

 only alternate centra supporting arches in this case. In Elasmobranchs 

 still another arrangement of arches and centra obtains. 



In Sijjhostoma it is stated by McMurrich that a distinct fibrillation 

 takes place in the fin-folds of the embryos, "which, increasing, leads to 

 the formation of horny rays,"* A similar remark applies to the devel- 

 opment of the rays of Hippocam-pus in the pectoral, anal, and dorsal fins. 

 McMurrich's statement, however, that the anal of Siphosioma aborts 

 after developing embryonic rays, must be denied, since the anal is 

 present in adults of both sexes of specimens of the same species which 

 have been examined by the writer. The interspinous cartilages are 

 also much more complex in the embryo of Hippocampus than in Sipho- 

 stonia of nearly the same age as those figured by McMurrich. The erec- 

 tores and depressores spina? muscles terminate as tendons which are as 

 long as the elements which I have elsewhere called hasiradial cartilages.] 

 These basiradial elements are cartilaginous, and each one at its distal 

 end supports a small cartilaginous nodule or actinophore, which is the 

 representative of a basilar interneural piece. Consequently the carti- 

 laginous bars, the whole of which McMurrich calls interspinous, are only 



* On the Osteology iind Development of Syngnatliii.s peckianus (Storer). Quar. 

 .Tourn. Mic. Sci., XXIII, pp. 623-650, 188:5. 



tA contribution to the development and morphology of the Lophobranchiates 

 (Hippocampui! antiquoruvi, the sea-horse). Bull. U. S. Fish Com., I, 1881, pp. 191-1 9U. 

 I plate. 



