MR. T. W. BRIDGE ON THE OSTEOLOGY OF POLYODON FOLIUM. 
727 
bars entirely devoid of any ossific centres, and, as is the case with no other Vertebrate 
animal, the parachordals are permanently distinct from one another, the notochord lying 
between them in a stratum of connective tissue. In Proteus, and also in Menopoma, 
the proximal segment of the hyoid arch persists as a large and distinct hyomandibular 
cartilage, and is especially large and attached by ligament to the periotic capsule in 
the former, so that the distal part of the arch is directly connected with the skull by 
its own proper proximal segment. In addition, there may be mentioned the absence 
of an opercular fold for the protection of the gills in all Perennibranchiata, with the 
exception of a rudiment in Siredon ; the persistence of external gills combined with 
the non-development of internal branchiae, the absence of eyelids, and the imperfect 
condition of the inter-auricular septum. Unless many of these features are due to 
the degradation of the more immediate predecessors of the recent Urodela, we must 
believe that the primitive common ancestor of the Urodela and the Anurous Amphibia 
was in many respects of a lower organization than any living Marsipobranch or Sela- 
chian. On the other hand, there are reasons for believing that the Perenuibranchiate 
Urodela owe much of their simplicity of structure to the fact that they are the 
descendants of Caducibranchiate forms which have undergone a certain amount of 
retrograde metamorphosis. The comparative ease with which the former under the 
influence of suitable natural or artificial conditions become Caducibranchiate, as shown 
by the interesting experiments of Fraulein Marie von Chauvin,* suggests that the 
change is nothing more than the assumption of a more immediate ancestral habit of 
life. As to the relation of the Ganoidei to this ancestral form, or to either of its two 
groups of descendants, two suggestions may be made. The close analogy between the 
Polyodontidse and the Anura in the nature of their cranial metamorphoses which I 
infer to exist from the existence and relations of the £i or bitar process ” in both groups, 
combined with the fundamental differences which exist both between the larval and 
the adult Anura and Urodela, is somewhat unexpected and inexplicable, on the view 
that the two latter are more closely related to one another than is either of them to 
any of the Ganoid Polyodon. On the other hand, is it possible that Polyodon may 
have been an offshoot from the first evolved Anura subsequently to the divergence 
from the Urodela, and that while the Ganoid has retained its piscine habit, the two 
other groups have been modified along parallel lines and adapted to an air-breathing 
life ? But an affirmative answer to this question involves the inadmissible conclusion 
that the phylogenetic relationship of the Anura to the Polyodontidae, and presumably 
therefore to all other Ganoids and to the Teleostei, is closer than that of the former 
to the Urodela. Such a conclusion is also negatived by many other considerations. 
The close correspondence which can be traced between the caudate and the non- 
caudate Amphibia in the arrangement of their generative and vascular systems, and 
m the structure of the vertebral column, pectoral and pelvic arches, and their limb- 
skeletons, is so marked that it can hardly be explained in any other way than by 
* Zeitschrift fur Wissenschaftliche Zoologie. Bd. xxvii., 187t>. 
5 A 
MDCCCLXXVIII. 
