667 



shoulder girdle. This idea was first expressed by C. W. Andrews 

 (Geol. Mag., Dec. IV, Vol. II, p. 81, Febr. 1895). He states that in 

 Keraterpeton Galvani Huxley, the paroccipital plates form a 

 long posterior process, which inK. crassum Fritsch are said to 

 be free, articulating with the paroccipital plates, and he makes the 

 following remarks in a footnote: "It may be pointed out that these 

 elements (epiotics [paroccipital plate]) occupy the position of the post- 

 temporals of fishes and may possibly indicate the former occurrence 

 among the Stegocephali of a connection between the posterior region 

 of the cranium and the pectoral girdle." Gegenbaur expresses the 

 same opinion in a footnote p. 17 of his paper. Since i have definitly 

 shown that the paroccipital plates and probably also their posterior pro- 

 cesses are homologous to the supratemporal and posttemporals (supra- 

 scapular) of Teleostomi, this opinion receives fresh support. 



The abdominal ossicles. 



In this connection I should like to mention the fine series of scale- 

 like ossicles which cover the ventral surface of the Stegocephalis they 

 form a series of transvers bands; each band consisting of numerous 

 ossicles. The fate of these elements we can follow in the Reptilia. 

 In the Permian Palaeohatteria and Kadaliosaurus we find that the 

 ossicles are reduced in number, but somewhat enlarged in size. In 

 the Ichthyosauria a still greater reduction has taken place; only five 

 bones being present, one median and two lateral ones on either side, 

 united to each other. In the Plesiosauria we find similar condition, 

 but the abdominal ossicles are very much stronger. In the Spheno- 

 dontidae, Homoeosauridae, Rhynchosauridae, Pleurosauridae, Saphaeo- 

 sauridae (Sauranodontidae) , the Pterosauria and Megalosauria, we 

 find only a median piece and two lateral ones. In the Crocodilia the 

 elements are in a rudimentary condition; there are two slender ele- 

 ments on each side, the median one being separated into two. It is 

 on these elements, that the middle and posterior portion of the plastron 

 developed. The plastron is a product of abdominal ossicles + ventral 

 dermal armour. 



The pelvis. 



In the Stegocephali we find for the first time among vertebrates, 

 the pelvis attached to the vertebral column. There is one sacral 

 vertebra, which is connected with the^ ilium by a rib; but the con- 

 nection is only a lorse one. In none of the Teleostomi is the pelvis 

 connected with the vertebral column. Ilium and ischium are always 

 ossified in the Stegocephali, the pubis only becomes ossified in the 



