[43] THE EVOLUTION OF CHE FINS OF FISHES. 1023 



erly the last portion of the dorsal, which is differentiated into three 

 portions. There seem to be no anal rays dev^iloped. 



The development of the Goose-fish or Lo2)hius, according to the ac- 

 count given by A. Agassiz,* is one of the most extraordinary yet known 

 to embryologists, and throws a great deal of light upon some of the 

 problems which this study has suggested, especially in relation to certain 

 questions which arise in the consideration of the genesis of the ventral 

 or pelvic fins and their genetic collocation with certain mesoblastic seg- 

 ments posterior to the pectorals, and far behind those with which they 

 are in apparent relation in the adult. 



In the youngest stage figured by Agassiz the pelvic fin arises as a 

 flat, lobate fold, some distance behind the pectoral fold. Fig. 1, PI. XI, 

 of nearly the same form and nearly synchronously with it. In the next 

 stage, Fig. 2, it is represented as arising below the base of the now 

 widened pectoral, as an elongate, spatulate process, arising from the 

 side of the yelk sack, with a blunted, somewhat swollen tip. In the next 

 stage it has become somewhat longer, more swollen at the tip, and its 

 base more advanced, Fig. 3. As development proceeds its base ad- 

 vances a little more in front of the point of insertion of the base of the 

 pectoral. The single first ray of the pelvic fin then buds out a second 

 ray near its base, and the first ray, which is now more elongate, be- 

 comes somewhat geniculate and swollen in the vicinity of the bend. 

 Its rays by this time. Figs. 4 and 5, far surpass in length those of the 

 pectoral, which are now only begin ning to develop. The rays of the 

 pelvic fin now continue to precociously lengthen until the outermost one 

 is nearly twice the length of the animal from the snout to the end of 

 the tail, the total length being 30™™. The bases of the pelvic fins are 

 now clearly in advance of the bases of the pectorals. 



This change in the position of the bases of the pelvic fins, if we keep 

 in mind the successive events which are mentioned above, would seem 

 to be due to an actual shiftims- forward of these fins from their original 

 position in close homonomous relation with mesoblastic segments be- 

 hind the pectorals. The jugular position thus assumed by the pelvic 

 limbs of LopMus during its early stages might readily affect their final 

 relations with the spinal nerves, and become innervated by pairs ante- 

 rior to those which appertain to the segments in direct relation with its 

 first rudiment when that was still behind the pectoral. HaswelPst 

 foot-note, in his paper on the Elasmobranch skeleton may, therefore, 

 lose its force if the preceding view is well founded. He says : " I 

 have ventured (On the structure of the paired fins of CeratoduSy Proc. 

 Linn. Soc. N. S. W., Yol. VII, p. 10) to make the very obvious sug- 

 gestion that the derivation of the pectoral and pelvic plexuses from 

 a number of spinal nerves was a strong piece of evidence in favor of 



*0n the young stages of Osseous Fishes. Part III. Proc. Am. Acad. Arts and Sci. 

 XVII, 1882, pp. 280, 285, Pis. XVI, XVII. 



t Proc. Liun. Soc. New South Wales, JX, 1884, p. 82. 



