On the Anatomy and Phyldgenetic Position of Polypterus. 407 



the saccular protuberance of the skull (fig. 24). In the 

 youngest specimen the bone forming in this region showed a tendency 

 to become separate though it never does so in Polypterus. However 

 in exactly the same spot arises the stapes or operculum 

 in theUrodeles as a small nucleus of cartilage belonging to the 

 auditory capsule. One may imagine the stapes of Urodeles arising in 

 some ancestral form as an ossification of the auditory wall, comparable 

 with the ectethmoid or with the pterotic of Teleostei, in the articular 

 region of the pharyngobraiichial. This bone became more independent 

 by gradual stages and replaced the pharyngobranchial itself. In the 

 Stegocephali it is occasionally massive. The pharyngobranchials became 

 lost and the stapes now developes as a separate nucleus. 



On certain of the dermal bones. 



The parietal must be regarded as a squamosoparietal (v. Wijhe). 

 The small mucous bones at the edge of the preoperciilum , including 

 the spiraculars with that attached to the hyomandibular, as also the 

 postparietals of Osteolepis and the small cheek bones of Lepidosteus 

 are interesting as evidences of the method of formation of cranial bones 

 but have no value for comparison with higher animals. That there 

 are 6 supratemporals in Pohjpterus may be primitive but their position 

 is not primitive. The typical number must be considered two. In 

 Osteolepis there are three, two paired and one median unpaired. In 

 Amia we find the typical two. Evidence of the secondary position of 

 these bones in Folypteriis is seen in tlie extension of the "squamosal" 

 process backwards under them. This squamosal process from the 

 squamosoparietal is curious. It runs down, that is, internally and 

 somewhat outwards and closely apposes itself, in the young specimens 

 to the supraoccipital transverse cartilage and the exoccipital and also 

 to the opisthotic. At this spot the bone so resembles bone arising 

 perichondrially that from the sections one could not determine, which 

 it was. In old specimens the anchylosis in this region is complete. 

 The projection then passes upwards and outwards and to it one of 

 the stoutest tendons of the neck muscles is attached while it is also 

 connected by ligament with antero-internal process from the posttem- 

 poral scale (figs. 13, 23). 



The posttemporal scale where it appears at the surface is oval 

 with the long axis directed towards the pectoral fin. It shows the 

 granulated appearance of the rest of the cranial bones, that is, this 

 portion is essentially superficial. It is supported, mainly by the opi- 



