GREGORY: FISH SKULLS 91 



It was the aim of the earlier naturalists, including Cuvier and Owen, to identify in the 

 fish skull those bony elements which appear to be homologous with corresponding bones 

 in the skull of man. As noted by Williston (1914), in the adult human skull there are but 

 twenty-eight bones (including the auditory ossicles), the primitive reptile has seventy-two 

 separate bones in its skull, while in the oldest ganoid fishes there were more than one 

 hundred and fifty. Consequently the supply of names from the human skull was soon 

 used up and new names had to be invented for the remaining elements. Fish and man are 

 very far apart, however, and in man the skull is so greatly distorted by the balloon-like 

 expansion of the brain, by the reduction of the jaws and their firm union with the skull, 

 that the equivalence of certain elements, such as the human squamosal, long remained in 

 doubt. 



Even in the times of Cuvier and of Owen comparison of the elements of the fish skull 

 with the human type would have been still more difficult had it not been for the existence 

 of living reptilian forms such as the crocodile, which serve in a measure to bridge the struc- 

 tural gap between the fish skull and that of man. Moreover since the days of the pioneers 

 in skull morphology several generations of palaeontologists have brought to light a whole 

 series of fossil forms which enable us to follow step by step the transformation of the skull 

 as a whole and the history of nearly every one of its elements. The present leaders in this 

 line of work are Professor D. M. S. Watson of London and Professor Eric A:son Stensio 

 of Upsala. Watson has adopted a system of nomenclature of the elements of the fish 

 skull which is based primarily on the names now commonly used for the corresponding 

 elements in the skull of the oldest known amphibians and reptiles. Stensio (1921, 1925) 

 follows the same general method, but he has shown (e.g. 1921, p. 139) that certain of the 

 bones present somewhat different combinations in diflFerent groups of Palaeozoic fishes; it 

 is also known that in the oldest fish skulls there are several elements, including the entire 

 opercular series, which have been dropped out even in the earliest known tetrapods. In 

 this country Dr. Starks (1926fl) has demonstrated the difficulties in homologizing especially 

 certain rather variable bones of the ethmoid region in diflFerent orders of teleosts; he has 

 also described the various combinations and replacements of compound bones of endosteal 

 or ectosteal origin, which further complicate the problem. Finally, despite all the work 

 already done the gaps between known crossopterygians and palaeoniscids, between known 

 palaeoniscids and holosteans, leave a few uncertainties as to the exact origin of several 

 bones of the preorbltal region of teleosts. Hence any system that may be used now must 

 be more or less tentative and it must also be more or less of a compromise if it is to be 

 practicable and serviceable. 



For many years past I too have been interested in this question of the nomenclature 

 of the skull elements of the lower vertebrates and have made repeated comparisons of skull 

 patterns in the series from fish to man, especially endeavoring to equate the leading systems 

 of nomenclature now in use in Europe and America (Gregory, 1917). In the American 

 school of ichthyology as developed by Gill, Cope, Jordan, Evermann and Starks, the main 

 object of the nomenclature of the fish skull was not so much to identify the fish-skull ele- 

 ments with those of man as to label the bones with appropriate names suggesting position 

 in relation to certain landmarks. In selecting names for the elements of the skull I have 

 tried to adopt those which are the most widely used, unless such names clearly imply incor- 

 rect homologies with the tetrapod skull. For instance, in common with Stensio and Watson, 



