360 Mr. A. S. Woodward on the 



unknown, but may be fused with the short ornamented 

 rostrum, which curves inwards to the mouth below and seems 

 to bear minute teeth. The mandible, as before, is shown to 

 be short and deep, the dentary {d.) forming by far the greater 

 part of the ramus. A fragment within the jaws may possibly 

 be part of the lower dental plate, but it is unusually pitted 

 and of doubtful nature. The operculum (op.) and the curved 

 anterior border of the preoperculum {p. op.) are also seen. 



The evidence afforded by the two skulls now described is 

 thus very suggestive, but not quite conclusive as to the 

 relationships of the dental plates named Plethodus. It is 

 still necessary to demonstrate by microscopical sections that 

 the plates in this small form of skull are truly Plethodus. It 

 is also essential to obtain a clearer view of the basicranial 

 axis before definitely deciding upon the homologies of the 

 upper plate. Meantime, however, it seems almost certain 

 that the concave plate of Plethodus was part of the upper 

 dentition fix(d to the basicranial axis, while the convex plate 

 belonged to the lower dentition and was supported by the 

 hyoid apparatus. The upper plate may have been ])art of 

 the parasphenoid — an arrangement common among fishes — 

 or it may have belonged to a much-extended vomer like that 

 of the extinct Pycnodonts. 



There is only one difficulty in regarding the concave plate 

 of P. expansus as homologous with tlie upper plate in the 

 small form of skull now described, namely that the aboral 

 lace is a little diflferent. The median ridge in the known 

 specimens of P. expansus does not definitely rise into a 

 vertical plate, though this may have been comparatively 

 fragile and readily broken away ; moreover, in the larger 

 plate the strongest lines in the fine reticulation on either side 

 of the median ridge are directed transverselv, while those in 

 the original of PI. XIV. fig. 2 are longitudinal. It is, 

 however, possible that accident in the first case and specific 

 difference in the other may account for the discrepancy. If 

 the homology prove to be well founded, it is likely that the 

 end of the plate connected with the ascending bar of bone 

 (PI. XIV. fig. 2, /y) is posterior. 



Finally, assuming that the new type of skull now described 

 does really represent the geims Plethodus^ it is still difficult 

 to hazard a suggestion as to the precise affinities of the fish. 

 The structure of the mandible shows that it is not a Pycno- 

 dont, while the a[)parcnt discovery of a hyoid or lingual 

 dentition opposed to the upper dentition of the mouth adds a 

 feature not previously known among Mcsozoic fishes. The 

 dental arrangement, indeed, is most closely similar to that of 



