Para sphenoid of a Paheoniscoid. 431 



because the actual basisphenoid is undeilaid by a posterior 

 broadened extension of the anterior shaft, which shaft and 

 extension we must recognize as the parasphenoid on account 

 of the dentate surface ah-eady described, and also by analogy 

 with the structures described above for Fishes, Amphibia, 

 and Rt-ptiles. 



I think the evidence may be accepted as conclusive that 

 specimens A, B, and C are actually compound bones built up 

 of a posterior basisphenoid and an anterior shaft-like para- 

 sphenoid, which, however, has a posterior extension under- 

 lying the basisphenoid and fused witli it ; and so we arrive 

 at the homology existing between the above specimens and 

 the similar bone in such types as MegalichthySj Loxomma, 

 Ophthalmosaurus, and the snake. 



Oidy one other point remains to be dealt with, and that is 

 a consideration of the anterior lateral processes of the present 

 specimens, which exhibit definite articulating facets. The 

 bones with which these processes articulated are unfortu- 

 nately not sufficiently preserved to allow of their characters 

 being determined, but there is little doubt but that they 

 are the pterygoids. 



These anterior lateral processes have an interesting 

 bearing on the question of the autostylic and hyostylic modes 

 of suspension of the jaws. 



Parker (8) and Swinnerton (9) have shown that Lepi- 

 dostexis and Amia possess siniilar processes from the trabecular 

 which articulate with a process (the " pedicle "") of the 

 palato-quadrate bar. Swinnerton discusses fully the import- 

 ance of this connexion between the bnsis cranii and the 

 palato-quadrate bar. Dealing with the Teleostomi. he shows 

 that there has been in the Teleosts a gradual reduction in 

 the metapterygoid legion of the palato-quadrate bar, con- 

 currently with an increase in the importance of the hyo- 

 mandibular. He also notes that in those lowly forms in 

 which the metapterygoid region is large, a pedicle is sent off 

 from its upper border towards the trabeculse, and the trabecul?e 

 have on each side a projecting process to meet this pedicle. 



Tlie evidence, Swinnerton says, " suggests that at some 

 former time an actual connection or articulation must have 

 existed between the pedicle [of the palato-qua(h-ate bar] and 

 the trabeculae." As already noted, Parker indicates that 

 such a condition exists in Lepidosteus, in which at all stages 

 from the embryo, two-thirds of an inch long, up to the adult 

 condition, there is a strongly developed pedicle forming an 

 articulation with the " basipterygoid " process. He says : 

 " the proximal part of each trabecula has developed an 



30* 



