728 SNOUT OF FCETUS OF MONOTREMES, 



which represents a section through the snout of a young Ornitho- 

 rhynchus of precisely the same stage of development as that 

 illustrated by Model ii. The section there figured is somewhat 

 posterior of the region represented by Model ii. 



These osseous prevomerine splints ( = paired rudiments of dumb- 

 bell bone) extend in the specimen itself from about 1 mm. behind 

 the hinder extremity of the true palatine process of the premaxilla, 

 backwards for a distance of 2 '6 mm. They begin in front in the 

 plane of the anterior end of the capsule of Jacobson's organ 

 about 1-89 mm. anterior to the anterior margin of Stenson's duct, 

 and end about 49 mm. behind its posterior margin. Here their 

 posterior extremities are slightly overlapped by the anterior 

 margin of the transverse cartilaginous lamina which corresponds 

 to that visible in Model i. (fig. 3, s.p.c). This lamina is the same 

 as that transverse strip of cartilage figured in the adult between 

 the dumb-bell bone and the osseous maxillary palate in fig. 17 in 

 the paper in the Macleay Volume already cited (6). It is also 

 shown in my figs. 8 and 9 ^^nf." in my former paper on the 

 dumb-bell bone (loc. cit.), and the relation to the bone is there 

 further indicated by the line "d" in fig. 1 of the same paper (lo). 

 But whereas in the adult I have shown the dumb-bell bone to 

 extend backwards into the roof of the internasal aperture there 

 described and depicted, here, in this early stage (specimen D, 

 Model ii.), the posterior extremities (not included in the model) 

 of the paired prevomerine elements as yet fall short of the anterior 

 margin of the aperture by about 0*49 mm. 



From this description and from the figs, of Model ii., it is 

 evident that the prevomerine dumb-bell arises autogenously during 

 the interval in development between the stages illustrated by 

 Models i. and ii., and is from the beginning separated by a wide 

 interval from the true palatine process of the premaxilla. The 

 latter has, indeed, obviously been receding in an anterior direction 

 ever since the stage of Model i., in which there was no trace of 

 the prevomerine elements. There cannot, therefore, at any time 

 have been any connection or continuity between the prevomers 

 and the true palatine processes of the premaxillee. This fact 



