Phylogeny of the Teleoslomi. 31:1 



plate. Paired postfrontals and suprateniporals are well 

 developed, whilst anteriorly a median ethmoid separates the 

 prromaxillaries. A single large bone on the cheek which 

 sends forward a process below the orbits represents the 

 snb- and postorbitals, and may include the maxillary also. 

 The opercular bones are represented by the operculum only. 

 The nostrils are lateral, between praimaxillary and ethmoid. 

 Gular plates and branchiostegal rays are apparently wanting. 

 In the arrangement of the bones of the cranial roof Coccosteus 

 is almost a typical Crossopterygian, and the arrangement of 

 the supports of the dorsal fin in two regular series and tlie 

 structure of the ventral fin, which appears to be essentially 

 similar to that of Foli/pterus*, cannot be said to negative this 

 view. 



A comparison of Coccosteus with Pterichthys shows the 

 following important points of agreeaient :: — ^ 



(1) The anterior part of tlie trunk is enclosed in an armour 



of bony plates which are not united to those of the 

 head, so tliat tlie latter is freely movable. 



(2) There is a single dorsal fin which is membranous. 



(3) There is a single opercular bone f and a pitted pineal 



bone. 



(4) The dermal armour \ is in hoth cases composed of dense 



bone loith a cancellated structure in its thicker portions, 

 with an outer layer of ganoine, ivith a tuherculated 

 surface, and ivith open grooves for the sensory canals. 



(5) The arrangement of the bones of the head, but espe- 



cially that of the dermal plates of the body, can 

 easily be reduced to a common plan. 



In the skull of Pterichthys we recognize posteriorly tlie 

 three dermo-occipitals, the suprateniporals, and the operculum 

 of Coccosteus, whilst anteriorly the median ethmoid and 

 laterally the large suborbital plate are still in the same relative 

 positions. The prgemaxillaries are now eiitirely on the lower 

 surface, but, as in Coccosteus, they seem to border the nostrils. 

 The orbits have approached each other until they are only 

 separated by the pineal plate. The postfrontal is fused with 

 the suborbital. 



If Jfekel be correct in regarding Ilomosteus as intermediate 



* It is noteworthy tliat Coccosteus resembles Polijpterus in the position 

 of the nostrils also. 



t In both cases this bone has been interpreted by some authorities as 

 other than opercular, so that it would bo perhaps better to say " there is in 

 both a similarly placed bone which ma}' be regarded as an operculum." 



X See Smith Woodward, ' Vertebrate Pakeontology,' p. 12 (181)8), and 

 Cat. Foss.Fish. ii. p. xix (1891). 



