GREGORY: FISH SKULLS 389 



seaweed. If separated from its seaweed, a Pterophryne can develop a surprising burst of 

 speed, wriggling its body and using all its fins to overtake its floating home. 



As their body has tended to assume the globose form of the puffer or porcupine-fish 

 type, the spinous dorsal has entirely lost its locomotor functions, while the soft dorsal tends 

 to be paired with the anal in a wig-wagging motion. Their highly cryptic coloration among 

 the sargassum weed seems to enable them to stalk successfully the small crustaceans and 

 other small animals and protects them from larger fishes that hunt in the weed. 



From all this we draw the conclusion that the marked peculiarities of the skull and 

 body-form in the higher pediculates were conditioned by the following primary adaptations: 



(1) the development of an expanded oro-pharyngeal chamber for engulfing a large prey 

 or a large amount of food; 



(2) the posterior spreading of the branchiostegal membrane, the closure of the normal 

 opercular slit and the migration of the lower part of this slit to a point behind and above the 

 pectoral fin — all correlated with the reduction of the gills, the development of the power 

 of inflating the throat and stomach with air or water, as well as the power of ingesting 

 large prey; 



(3) the elongation and specialization of two of the pectoral pterygials so as to enable 

 the pectoral paddles to reach around below or above the swollen throat and abdomen; 



(4) the change of function of the enlarged first ray of the spinous dorsal into a lure. 

 Probably the initial step was the habit of living among seaweed, the acquisition of dappled 

 color of skin and the development of excrescences or tags of skin on the body and anterior 

 dorsal fin. More or less directly progressive stages in this transformation may be seen in 

 Antennarius {" Ckironectes") unipennis, A. scaber, A. lophotes, A. mummifer. 



(5) As to adaptive radiation into different habitats, I infer that the primitive habitat 

 was in the kelp along rocky shores; that some of the antennariids clung to the Gulf-weed 

 when it broke loose from the shore and was carried far out into the Sargasso Sea; that, as 

 this weed eventually sinks, some of them became pelagic and free-swimming and thus gave 

 rise to the various families of ceratioids; that in another direction, some of the early anten- 

 nariids, becoming benthonic and with ever greater mouths, gave rise to the lophiids, while 

 others through ChaunaxAWe forms passed into the sea-bats (Onchocephalidie). 



After the foregoing summary we are perhaps in a more favorable position to attempt an 

 evolutionary interpretation of the architecture of the skull of pediculates. The following 

 sources and material have been used: 



(1) various dried skeletons of Lophius^ Antennarius, Pterophryne and Ogcocephalus, in 

 the collections of this Museum; 



(2) the fundamental article by Tate Regan on the classification of the order Pediculati 

 (1912/) and his memoir on the Ceratoidea (1926); 



(3) the ceratioids in the collections made by Dr. William Beebe for the New York 

 Zoological Society. These will be more fully described by him in subsequent publications, 

 but while I was enjoying the hospitality of his laboratory in Bermuda he very generously 

 invited me to study and sketch any of the deep-sea material and to use as much of it as I 

 cared to for the present paper. 



Antennariids. — The skulls of Antennarius and its allies {Histrio, Pterophryne, etc.) are 

 on the whole perhaps the most central and least specialized of the pediculates above the 

 grade of the batrachoids and well reflect the influence of the primary adaptations discussed 



