386 TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



described characters of the skull of Opsanus are no older than the inward-turning of the 

 teeth of the mandible. In other words, such "habitus" features may have been acquired 

 in a relatively short (geologic) time and hence the relationships with other fishes with very 

 different skull patterns may be less remote than their appearance would indicate. This may 

 be one of the reasons why some systematists like Boulenger have found skull characters 

 relatively unstable and have had to seek for evidences of more remote genetic relationships 

 in the relations of the ribs to the vertebral column and in the form and relations of the parts 

 of the bones of the pectoral girdle and fins. 



There is no reasonable doubt, for example, that the toad-fishes are related to the 

 pediculates but they differ much in the construction of the ethmoid region {cj. Starks, 1926a, 

 p. 306). Another fact that tends to disguise relationship between fishes of different habitus 

 is that a given bone may change its size or one of its contacts with surrounding bone either 

 through what seems to be a growth or reduction of its own or through a more widely spread 

 growth or reduction that affects whole regions rather than particular bones. For instance, 

 in the toad-fish the marked transverse growth and depression of the head as a whole appar- 

 ently caused the hyomandibulars to become widely divergent, broadened transversely and 

 shortened vertically. The preoperculars shared the same distortion but the operculars 

 immediately behind them have become quite small and now serve chiefly as a support for 

 two aggressively large spikes, which not improbably are able to inflict severe wounds. 



Again, the whole region of the primary upper jaw and its supporting parts in the toad- 

 fish were enlarged, while at the same time the premaxillae and maxillae dwindled to slender 

 bars. This perhaps exemplifies the old principle of "compensatory growth." 



In any case the fact remains that among the forms available for examination no one 

 appears to give the clue as to the precise point of origin of the batrachoid-pediculate stock 

 from any of the percomorph fishes. Tate Regan (1912/, p. 279, Pediculati) even suggests 

 the possibility that this group may have come off from a pre-percomorph stock in which 

 the hypurals had the relations to the centra which are exemplified in the Salmopercae. 

 Probably this means at most that the batrachoid-pediculate branch (Lophius being already 

 present in the Upper Eocene of Monte Bolca) may have begun to diverge from the acanthopt 

 stock before the beginning of the modern families of percomorphs. 



That the batrachoids are related to the pediculates is well attested by their reasonably 

 close approach to that peculiar group in the characters of the pectoral fins, vertebrae, 

 hypurals, as well as in many significant skull characters. While in general more primitive 

 than the pediculates, they are more specialized in the loss of the mesethmoids and of the 

 epiotics, or possibly in the fusion of the latter with the parietals (Tate Regan, 1912/, pp. 

 279-280). 



Mr. G. Allan Frost (1930^, p. 623) states that the otoliths of the order Haplodoci 

 show no resemblance to those of the Pediculati but in their general features they resemble 

 in a striking manner the otoliths of the family Macruridae of the order Anacanthini, dif- 

 fering, however, in minor details. 



Pediculati (Anglers, Sea-mice, Sea-bats) 



Before discussing the evolution of the skull of the pediculates it is well to consider briefly 

 the evolution of their general body-form. The name Pediculati is given of course in refer- 

 ence to the peculiar construction of the pectoral fin, which can be turned downward, back- 



