178 PEOP. W. K. PAEKEE ON THE MOEPHOLOGY 



adult skulls have much in common ; but this is move massive (Pis. XVI. & XVII.) and 

 longer. 



In these full-gilled larvae the occipital arch (PI. XVII. figs. 1-3) is well ossified, and 

 the ossification on each side runs continuously over the inner part of each auditory sac, 

 so that the exoccipitals and prootics (e.o,pro) are not separate. There is a distinct and 

 widish tract of super- and basioccipital cartilage (figs. 1, 2), and a solid facet of cartilage 

 for each condyle (oc.c). These oval elevations look backwards and inwards. The inter- 

 space between them is square and large ; it is filled by a rudimentary vertebra, whose 

 centrum has coalesced with the centrum of the " atlas," between its large pedunculated 

 " prsezygapopbyses." 



In one specimen (figs. 1 & 2) there was scarcely more conjugating cartilage in the basi- 

 occipital region than is seen in the adult Proteus and Menobranchus s, for the parachordals 

 (iv) had retreated already from the trabec ulae, whose hind part (its parachordal tract) 

 had been already absorbed (figs. 1, 2, tr, iv, nc) ; but in the other larva (of the same 

 size) there was a U-shaped posterior basicranial fontanelle, traversed by the thick 

 notochord, and bounded in front by the thick broadish trabecular band (fig. 3, tr, nc, 

 p.bc.f, iv). This is the band which is persistent in the Salamander (PI. XVI. fig. 4), but 

 which so soon vanishes in this type, but not always at the same time in different 

 individuals. 



Laterally the trabecular were confluent with the parachordal cartilages and also with 

 the cartilaginous regions of the ear-sacs ; beyond that point they had grown into side 

 walls to the skull, which were of great height in reality, but bulged considerably. 



In the ripe Salamander's larvae (PI. XV.) there was no appearance of bony deposit in 

 the side walls ; but here, from the optic foramen (2) to a short distance from the nasal 

 region, the whole wall is solid bone (figs. 1 & 2, sp.e). Yet in front of these tracts the 

 whole of the structures represent a very early condition in other types (e. g. Siredon, 

 Salamandra). 



Three pairs of splints have not yet appeared, namely the nasals, ectoethmoid, and 

 maxillaries ; this stage thus corresponds with the adult Proteus and Menobranchus *. 



Three pairs of cartilaginous growths are seen here, of which only one pair occurs in 

 the first of those " Perennibranchs ;" and one tract seen in them is wanting here ; for the 

 long internasal part of the trabecular has no conjugatory tract, as in Proteus and 

 Menobranchus', but what they lack these have, namely, a large flat lobe outside the 

 end of each rod (c.tr) ; these give the form of a hatchet to the fore part of the trabecular. 



These larvae correspond with those Perennibranchs in having a distinct " antorbital " 

 or ethmopalatine cartilage (e.pa) on each side ; it is, however, shorter and broader than in 

 them (Huxley, op. cit. pi. xxx. fig. 1, A.o, and PL XVII. figs. 1 & 2, e.pa). But they 

 have acquired a new structure of intense importance ; this is a sickle-shaped olfactory 

 cartilage, or nasal roof (na) f , which is absent in Proteus and imperfect in Menobranchus. 



* For the figures of these types, see Huxley, P. Z. S. 1874, pis. xxix. xxx., and my recent paper in the Phil. 

 Trans. " On the Skull of the Urodeles." 



t After Prof. Huxley's paper was published, he expressed to me his doubt of the distinctness, in any case, of this, 

 the foremost of the " paraneural " cartilages : the figures here given may serve to expel that doubt. The chondro- 

 cranium of Menobranchus, as figured by him, has much less basal cartilage in it than it had at an earlier period. 



