GAR-PIKES, OLD AND YOUNG. 



193 



middle, and dissection shows that the spinal axis is continued back- 

 ward and upward as a cartilaginous rod,' terminating at the upper 

 border, just under the hinder pair oi fulcra, and at the point where the 

 filament was attached. The raj^s are all attached to the lower border 

 of the spine ; and there is only a lower lobe of the tail. 



A similar structure exists in the tail of Amia, which Prof. Huxley 

 gives as an example of heterocercal tail. It seems better, however, to 

 discriminate between it and the previous stage, where the upper lobe 

 (filament) exists, and it may, therefore, provisionally be called the 

 masked heterocercal, or perhaps the pseudo-homocercal. 



Prof. Huxley has more recently given figures and descriptions of 

 the tail of embryo Teleosts (Gasterosteus), in which the structure is 

 nearly identical with that of the adult Amia and LepidGsteus.'^ 



A. Protocercal. First stage of Lepidosteus. Permanent in Am- 

 phioxus. Petromyzon. Lepidosiren, Polypterus. Also in some 

 ancient Ganoids, as Glyjjtolcemus. 



B and G. Heterocercal. In the sturgeons, and most sharlis, and 

 many mesozoic fossils. 



D. Not represented, so far as I know, among recent or fossil 

 fomiB. 



E. Masked heterocercal. In adult Amia and Lepidosteus. In the 

 embryo of many Teleosts. In Megalurus and some other 

 fossils of Mesozoic and more recent epochs. 



Fig. 10 Diagrams intended to illustrate the Cokkespondence of the Successive 

 Stages op Transformation of the Tail of Lepidosteus, with the Tails of Certain 

 Living Forms more and less generalized, and op Certain Fossils more and less 

 ancient. 



A, the first or rotocrra/ stage, where the end of the vertebral column (Fc) is horizontal and di- 

 vides the tail into upper and lower lobes nearly equal in size. B and Cthe heterocercal stage, 

 where the original tail is more or less elevated by the lower or infra-caudal lobe (/C), and De- 

 comes the filament (Fi), usually called the "upper lobe." In D the infra^caudal lobe is longer 

 than the filament, and in ^-the latter has wholly disappeared, and the tail assumes the last or 

 'masked heterocercal" condition. 



The same author concludes that in many adult Teleosts the poste- 

 rior end of the spine is more or less strongly bent up, although the tail 

 is outwardly nearly or quite symmetrical. 



1 This rod consists of the notochord, and a slender prolongation of the spinal cord, 

 surrounded by a cartilaginous sheath. 



2 The writer has found the same condition in newly-hatched catfishes (Ammrus), and 

 it has been observed in the embryo of a species of Cottus, by Mr. S. H. Gage, a student of 

 natural history at Cornell University. 



VOL. XI. 13 



