194 PROF. W. K. PARKER ON THE MORPHOLOGY 



lanceolate. The posterior nares (i.n) are in their usual relation to the vomers; but there 

 are no maxillaries outside, nor nasals and ectethmoids above. 



The two pairs of roof-boues (f,p), with the nasal processes of the premaxillarieSj are 

 adapted to each other by wedge-shaped imbrications ; the frontals reach the internasal 

 cartilage, and the parietals overlap the occipital belt slightly (fig. 1). 



The parasphenoid (fig. 2, pa.s) is oblong, with rounded corners, and is further irregular 

 by increasing somewhat from before backwards : the vomers underlie it a little in front; 

 it reaches nearly to the foramen magnum behind. 



The squamosals (sq) are nearly equal in their supratemporal and preopercular regions. 

 It will be seen that I have figured one and part of another vertebra. The first of these 

 is imperfect, having no arch ; in the other the arch has been cut away. This second 

 joint carries the skull on its long pedunculate " pre-zygapophyses," and between these 

 the curious elongated centrum of the intercalary (odontoid) vertebra is fixed. It is 

 ankylosed to the second centrum behind, and in front carries a bilobate mass, having a 

 form somewhat like dumb-bells. 



As this intercalary vertebra is inconstant, and as it may represent part of the occiput 

 in the Sauropsida and Mammalia, for their hypoglossal is a cranial nerve, but is 

 spinal in these low types, I shall describe this segment the more carefully. 



In these types the notochord shortens rapidly (figs. 4 & 6, tic), whilst in Seironota the 

 cephalic part is large, and partly segmented into two imperfect transient cranial centra. 

 But if the cephalostyle is absorbed in Spelerpes, the notochord makes up for it in the 

 intercalary segment. In the large semicircular notch between (not behind) the occipital 

 condyles a third pair of "parachordal" cartilages appear (besides the ordinary pair 

 and the parachordal region of the trabecular). These nuclei of cartilage become ossified, 

 and then coalesce by a bridge of bone; hence their dumb-bell shape. Meantime the long 

 tract of notochord has become ensheathed with bony substance, and its apex embraces 

 the bony stolon between the lateral lobes, and coalesces with it, its own end being 

 emarginate, and more developed below than above the transverse plate. 



Here we have the three normal elements of a vertebral centrum ; and this curious 

 structure is in front of the atlantean segment, a perfect and large vertebra, made to 

 carry the skull. Thus the human odontoid process, which is merely the ossified noto- 

 chordal core of the " atlas," enucleated from the symmetrical sides, and ankylosed to the 

 " axis," is here represented and foreshadowed by a distinct mcta-somatomic joint, with 

 middle and side elements combined. 



The inferior arches scarcely yield in interest to the skull itself, and show that these 

 types are certainly related to the lowest sort of Perennibranchs. 



The cartilaginous pith of each mandibular ramus is very solid, especially towards the 

 articular end ; that part rests in a delicate bony trough, the articulare (PI. XIX. fig. 9, 

 ar), and the outside is invested by the bony element (d). Where these bones meet on the 

 inside, just in front of the highest or coronoid part, there is a small dentigerous plate, the 

 splenial (sp). Here the splenial is very small ; in Proteus it also is so, and that 

 type has no hypohyal. Here it is only indicated by a groove crossing the inner face 

 of the bar below {c.hy, h.hij). The branchials are normal as to form, but abnormal as to 



