300 PROF. W. K. PAIIKEB ON THE STRUCTURE AND 



and these wings are direct upgrowths of the paired trabeculse, which grow, like new 

 shoots, from the under surface of the upturned parachordals. 



In front of the pituitary involution, there being no axial notochord, a new axis 

 appears ; this preaxial rod, the " iutertrabecula," is very large indeed in the second 

 stage in this type. It is seldom absent (as in the Ophidia), but in all forms that have 

 a projecting snout this bar, rod, or plate forms the axis of such foregrowth of the skull, 

 and may be equally hypertrophied in the Mammals (Cetacea) as in the Elasmobranchs 

 (Pristidse). 



The Crocodilia, as this stage shows, have inherited a compound nasal labyrinth — (a) the 

 valvular cartilage outside (in front), {h) the ethmoidal region, behind, (c) the proper 

 nasal roofs, and [d) the middle wall formed by the iutertrabecula ; these are all chon- 

 drified continuously, and we must go down to the Amphibians, Elasmobranchs, and 

 Marsipobranchs for a true interpretation of what is seen in these high Reptiles. 



In this stage the clefts are fast closing in, and the inner opening of the jiersistent 

 pouch, the first or tympanic cleft, is a mere crescentic slit, and corresponds to what is 

 permanent in some Batrachia. 



Instead of the ichthyic, large, symmetrical, perforated pharynx there is here a mere 

 funnel-shaped enlargement of the fore end of the oesophagus; and only three of the 

 visceral arches are developed at all, whilst only the first attains to its full size. 



AVith the total loss of branchial function there is an extraordinary amount of new 

 specialization ; and only by tracing out the early stages can a true interpretation of the 

 parts be made. 



In this stage the first and second visceral arches, if compared with those of the 

 Elasmobranchs, Chimajroids, and Urodeles, will be found to be normal, or nearly so. 



There are only two main segments on each side in the mandibular arch, viz. the 

 pterygo-quadrate and the articulo-Meckelian. The pier of this foremost arch has a huge 

 " otic process," and a rudimentary " pedicle " with two forks — one a small " ascending 

 process," and the other a small " pterygoid cartilage." 



The next arch has several normal branchial segments, like its counterpart, the hyoid 

 of Chimcera ; but it has also supernumerary segments like those seen in Acipenser and 

 in other Ganoids. As in many Selachians, its pharyngo-hyal or upper joint pushes 

 itself into the side of the auditory capsule; below this there is an epihyal and a 

 ceratohyal, the latter stopping short behind the hinge of the mandible, and becoming 

 solid continuously with the articular head of that part, its proper serial homologue. 

 Distally there is but a small region of basihyal cartilage continuous with that of the 

 third arch or fii'st branchial, proper. That arch has merely a small pair of cerato- 

 branchials. 



Passing on to the third and fourth stages, we find that the whole of the chondro- 

 cranium (with its visceral arches) has become Sauropsidan, and the investing bones, 

 which are now demonstrable, are in number and relation quite Crocodilian. 



