238 J. K. Thacher — Ventral Fins of Ganoids. 



The other Ganoids do not offer us much of interest in their ventral 

 fins, exhibiting as they do only modifications by concrescence, and 

 showing no advance either in the direction of the Dipnoans or the 

 Stapedifera. Figure 1, Plate II, represents the j^elvic girdle and rays 

 of Polyptervs. The fact that the two pubic parts abut on one 

 another without uniting, though fastened together by tough con- 

 nective tissiie, that there is no iliac process and that the whole struc- 

 ture lies in one horizontal plane should be stated. If this fin is 

 compared with that of Acipenser and Seaphirhynchus there can be 

 no doubt, I think, that the solid pubic pieces in Polypterus represent 

 the basal segments of rays, and therefore correspond to the orad 

 proximal parts formed of united rays in Polyodon and the Acipen- 

 serids. Some one or more of the four rays in Polypterus may be 

 composed of more than a single ray. This is not easy to determine, 

 but there is no doubt that they correspond to the middle and terminal 

 parts of the complete tri-segmented ray, and that the terminal seg- 

 mentations have disappeared as they have in the orad rays of the 

 Chondrostei, and so very generally in the orad rays of Elasmobranch 

 ventral and median fins. 



The fins of Lepidosteus and Amia ai"e not remarkable for much 

 beside their similarity. They are represented in PI. II, figs. 2 and 3. 

 Comparing them with the fins of Polypterus we see that they seem 

 to offer an extension of that same process of reduction which we 

 noticed in that fish. The parts lettered bd certainly are homologues 

 of one another in all three. But the middle and terminal segments 

 of the primitive ray, represented by the four well-developed rays of 

 Polypterus, are here reduced to the insignificant parts at the extremi- 

 ties of the main portions h d. The median one of these small pieces 

 is somewhat produced and ossified. The pubic parts overlap one 

 another but do not join, being merely fastened together by connective 

 tissue. And both in Lepidosteus and in Amia the overlap})ing was 

 in the same way, that is, the extremity of the right pulns was dorsad 

 of that of the left. There is no iliac process in either of these forms. 



We thus see that in these living Holostean Ganoids the only modi- 

 fication is the reduction in the number of the rays, (and this may be 

 only apparent) and the concrescence especially of the proximal seg- 

 ments. They offer us no advance toward the Dipnoan or Stapediferal 

 limb-skeleton. As to the light which they throw backward on the 

 earlier history, the separation of the pubic part of one side from that 

 of the other and the absence of an iliac process is to be noticed. 

 Otherwise they are uninteresting forms. 



