[79] THE EVOLUTION OF THE FINS OF FISHES. 1059 



exserted to a variable extent in tbe different forms, and by just so much 

 do they differ as regards the point where the ventral lobe of the caudal 

 grows out even in the embryo, so that traits appear very early in the 

 development of the caudal which serve as marks which characterize the 

 si:)ecies. The hypertrophy or local outgrowth of mesoblast to form the 

 lower lobe is variable in position to a very great extent, so that the 

 l)oiut where displacement due to growth occurs, also varies in position 

 in respect to the end of the chorda. This variation may be well seen 

 in the figures of embryos given by A. Agassiz and other embryologists. 



The development of the lower lobe of the caudal by the proliferation 

 of mesoblast and hypertrophy of a part of the hypaxial fin-fold has letl 

 to the final elision or suppression in some cases of the larval opisthure^ 

 due to a substitution of a secondary or permanent caudal so developed. 

 The localization of the energy of growth in advance of and below the 

 end of the chordal axis has brought about the abortion of the opisthure 

 or what must primitively have formed all of the tail of the ancestral 

 type. But the stimulus which led to the hypertrophy of some part of 

 the hypaxial median-fin-fold must have been the penultimate efficient 

 cause of the development of the opisthure, which becomes rejected or 

 elided by change of function or functional substitution. That stimulus 

 was applied locally to the hypaxial fold, and the hypertrophy which fol- 

 lowed was a natural result of use and effort, which very probably arose 

 from the continually-repeated efforts of the animal to rise to the surface 

 to get away from the bottom, and maintain itself right side up, as a re- 

 sult of which the growth of the ventral fold was stimulated and extended 

 ventrally. 



This view of the matter is strongly favored by the theory of the 

 median fins which has been defended throughout this essay. The con- 

 tinuous system of median rays found in a perfect condition of diphycercy 

 or the protopterygian stage would be available for a starting point, from 

 whence subsequent heterocercy might arise. The Chimseroid fishes real- 

 ize this condition, as was well stated by Balfour and Parker,* as Ibl- 

 lows : " The tail of Ghimwra appears to us to be simply a peculiar modi- 

 fication of the typical Elasmobranch heterocercal tail, in which the true 

 ventral lobe of the caudal fin may be recognized in the fin-fold imme- 

 diately in front of the filamentous portion of the tail. In the allied 

 genus Callorhynchus this feature is more distinct. The filamentous 

 portion of the tail of Ghimwra constitutes, according to the nomencla- 

 ture adopted above, the true dorsal lobe, and may be partially ijaralleled 

 in the filamentous dorsal lobe of the tail of the larval Lepidosteus.^^ 

 ******* 



How near these authors were to a clear comprehension of the true 

 nature of the tail of Ghimwra may be gathered from the foregoing ex- 

 tract, but I would take exception to their calling the tail of this form 

 heterocercal; that it is not, for tbe chordal axis has not suffered upward 



* Structure aud development of Lepidosteue. Philos. Trans., Pt. II, 1882, p. 408. 



