87] THE SKULL OF AM I URUS— KINDRED 87 



the anterior and lateral septa, but does not meet the pterotic lamella which 

 lines the lateral recess. It forms an osseous sheath on the anterior face of the 

 posterior septum and encloses the fenestra for the passage of the posterior 

 membranous canal. Below this fenestra it meets the exoccipital. The inner 

 end of the canal for the passage of the ramus laterahs accessorius of the facial 

 nerve and the jugular vein lies in a recess in the bone above the fenestra. In 

 median section this bone shows itself to be well ossified and thick, especially 

 in that region which forms the margin of the posterior fontanelle. Laterally, 

 the inner and outer lamellae are separated by cartilage. 



This is one of the best developed bones of the adult teleost craniimi and 

 usually has a long posteriorly projecting spine attached to its dorsal surface. 

 Cuvier recognized it in Perca and stated that, while it might possibly be 

 regarded as the interparietal, he preferred to describe it as the homologue of 

 the reptilian supraoccipital. Hallman (1837) figures it in Cyprinus and Perca, 

 but gives no detailed description. Parker (1872) was the first one to describe 

 the development of the bone in detail in the teleosts, as arising from perichon- 

 drial lamellae on the occipital arch and synotic tectum, between the parietals, 

 and touching the frontals anteriorly. In Salmo the margins of the bone are 

 rounded and there is not the spiculate serrate edge found in Amiurus. Sage- 

 mehl (1885) described the bone in the Characinidae briefly, and commented 

 upon the extent of the occipital spine, with the remark that from its relations 

 to the muscle, there was evidence that it had been developed from the fascia 

 between them and has secondarily fused with the main part of the bone 

 developed on the occipital arch. In his discussion of the bone in the Cyprini- 

 dae, he states an hypothesis for the development of the bone from the ganoid 

 condition, where it is wanting, by assuming that the occipital arch in the 

 teleosts, upon which the supraoccipital bone developed, arose from the assimila- 

 tion of the first vertebrae in a ganoid ancestor of the teleosts, and that there is 

 no evidence for regarding the bone as a derivative of the dermal plates in this 

 region of Polypterus, Acipenser, Lepidosteus and Amia. Loomis (1900) has 

 shown in the fossil ganoids of Kansas, that the supraoccipital is absent, 

 although present in the fossil teleosts. Zittel (1884, 1893) and Woodward (1898) 

 have described this same condition in the fossil Teleostomes. But in a previous 

 discussion, it was concluded that the vertebra at the anterior end of the 

 vertebral column were not serially homologous in the different groups and that 

 segments can be intercalated and excalated. So instead of regarding the 

 supraoccipital as the homologue of a neural process of the ganoids, it must be 

 assumed to be a new part which is intercalated as a new formation in the tele- 

 osts, but at what time this intercalation took place there are no fossil records. 

 The homologue of the supraoccipital plates of Polypterus is the spina occipitis. 

 This has been developed from connective tissue above the occipital arch and is 

 fused to the underlying supraoccipital ossification. In the Stegocephali 

 (Fritsch, 1883), there are a pair of supraoccipital plates corresponding to those 



