92 TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



I reject the name squamosal for the bone which is lateral to the parietal, and adopt for this 

 element the name pterotic, which is known to all ichthyologists. The superficial part of 

 the pterotic is very probably equivalent to the true supratemporal of the earliest tetrapods, 

 but as the name supratemporal has been applied to several different elements in this region 

 of the fish skull, I choose to treat supratemporal as a synonym of pterotic. 



In Table I the elements are classified according to their functional and topographic 

 relations with the olfactory, optic, otic and encephalic organs, the primary and secondary 

 jaws, hyoid and branchial regions. So far as practicable the distinction between endo- 

 cranial (deep, endosteal) and dermocranial (superficial, ectosteal) bones has been indicated, 

 but Starks has shown that in the cases of the ethmoid, sphenotic, pterotic, palatine, ptery- 

 goid and entopterygoid, we may be dealing either with an endocranial element or with a 

 "derm bone" or with various combinations and replacements of inner and outer elements — 

 all of which naturally make it impossible to avoid wrong implications of homology with 

 corresponding elements of tetrapods whenever the same name covers different combinations 

 in different groups. 



The main steps in the development of the nomenclature adopted in this paper may be 

 summarized as follows: 



(1) The history of such skull bones as are retained in man have been followed backward 

 to the oldest tetrapod stage. ^ 



(2) "Williston's law" ^ of the progressive reduction in the number of skull bones in 

 passing from lower to higher vertebrates has been critically examined and confirmed by 

 the work of many authors, including Stensio, Watson, Broom, Williston, Case, and von 

 Huene. 



(3) Such bones as do not survive into the mammalian stage but are known in the older 

 reptilian and amphibian stages (including the supratemporal, intertemporal, postorbital, 

 postfrontal, prevomers, ectethmoids, etc.) have been followed backward to the oldest tetra- 

 pod stage. 



(4) The history of all bones which have been named first in the higher stages has been 

 followed backward as far as possible. 



(5) The history of those extra bones of the fish skull which are present in the older 

 and tend to be lost in the later stages has been followed forward. 



(6) The elements of the oldest tetrapod skulls have been compared and, so far as 

 possible, equated with those of the oldest and most primitive fishes. 



(7) {a) "Tetrapod" names have been used for certain bones when the homologies 

 seem established beyond reasonable doubt (e.g. parietals, f rentals); {b) "Fish" names 

 have been retained when the homology with tetrapods is in doubt or when the appearance 

 and function are extremely different in the two groups (e.g. the quadrate of fish is homolo- 

 gous with the incus of mammals, but to call it incus would be to introduce a new element of 

 confusion in an already complicated nomenclature). 



Elements of the Preorbital Region 

 In certain Devonian crossopterygians, in which the skull is probably on the whole 

 much more primitive than that of teleosts, the dorsal and lateral surfaces of the rostrum 



' For a summary of the main results in this field, see Gregory, 1927, The Palseomorphology of the Human Head: Ten 

 Structural Stages from Fish to Man. Part I, Quart. Rev, Biol., H, No. 2, pp. 267-279; 1929, Part H, Idem., IV, No. 2, pp. 

 233-247. 



'This law is clearly formulated by S. W. Williston in his "Water Reptiles of the Past and Present," 1914, p. 3. 



