Phylogeny of the Teleostomi. 339 



Ciossopterygii, but the supposed homology of the lateral gulars 

 with the brauchiostegals is doubtful. As has been pointed out 

 above^ in the Pak«oniscida3 the gular plates and branchiostegal 

 rays are serially homologous, whereas the Crossopterygian 

 lateral gulars are plates developed between the principal gulars 

 and the mandibular rami. Moreover, whilst the Palaeoniscid 

 brauchiostegals are so imbricated that each overlaps the one 

 in front of it, the lateral gulars exhibit precisely the reverse 

 arrangement. Nevertheless, in the Devonian Palseoniscid 

 CAet^-o/e/'z's, as figured by Traquair *, the anterior branchio- 

 stegal extends forward between gular plate and lower jaw, 

 and this might be regarded as leading to the Crossopterygian 

 condition. 



In the Crossopterygii we see the development of the bone 

 which Boulenger has shown to be the representative of the 

 irquaniosal of higher Vertebrates. This is fused with the 

 prajoperculum in Polypterus, but coexists with it in several 

 extinct forms, and corresponds to the upper bone of the 

 po.storbital (as distinct from the circumorbital) series of the 

 Pala^oniscida?. The bone internal to it, which is the one 

 usuiilly called squamosal in fislies, is without doubt the true 

 suprateniporal t, and should be so named throughout the 

 Teleostomi, whether or no it includes a " pterotic " ossification 

 in certain Teleosts, whilst the series which lie posterior to the 

 parietals and true supratemporals might be termed dermo- 

 occipitals, thus avoiding confusion with the true supraoccipital. 



Many Crossopterygii have a pineal foramen, a feature as 

 yet undiscovered in any Chondrostei, and they must have 

 evolved in the Silurian from some primitive type belonging 

 to the latter order. 



DiPNEUSTI. 



The relations of the Dipneusti to the Crossopterygii have 

 been elucidated by DoUo % i'^ ^ convincing essay. He gives 

 good reasons for believing that Dipttrus is the most generalized 

 ot all Uipneusti, and that it has originated from a Crosso- 

 pterygian type closely allied to Holoptychius. It is only 

 necessary to add iiere that his views as to the specialized 

 character of the lobate paired fins receive additional con- 

 firmation from the demonstration of the primitive nature of 

 the non-lobate paired fins of the Chondrostei. 



* Ann. & Mag. Nat. Hist. (4) xv. 1875, p. 237. 



t This conclusion is not invalidated by the fact that Tohjpterus has no 

 suprateniporal, the bone so named by Eoulenger being- the " accessory 

 hyomaiidibular" of Traquair. 

 ' X Bull. See. Belg. Geol. ix. 1895, p. 79. 



