with the hyomandibula, it agrees better with the pharyngo- 

 suprahyal, which contacts the otic capsule below the lateral 

 head vein and in front of the hyomandibular nerve. The 

 palatine and hyomandibular branches in the tetrapods 

 usually emerge separately, suggesting the involvement of a 

 "lateral commissure." No such commissure is indicated m 

 development and indeed could not be present if its substance 

 is involved in the formation of the stapes. 



The dermal bones of the skull appear early in the course 

 of development, the tooth-bearing bones of the mouth first, 

 followed by those of the cranium. Canal bones are found 

 along the infraorbital canal, starting with the postorbital 

 and continuing through an infraorbital (jugal), and lacrimal. 

 Two small extrascapular canal bones lie along the posterior 

 margin of the roof. The bones of the roof arise deep in the 

 dermis and are never related to canal units. In Protoptenis 

 and Lepidosiren the posterior cranial roof lies below the ver- 

 tebral muscles; in Neoceratodus the muscle lies between this 

 roof and the endocranium. 



General observations on the choanate fishes 



Whereas the crossopterygian agrees in many details with 

 the tetrapod pattern, the dipnoan does not. Because the 

 living lungfish is less like the crossopterygian in its bone pat- 

 tern than its early fossil predecessors (Dipiems) were, it can 

 be assumed that a process of change had already begun 

 before the lungfish were first encountered in the fossil record. 

 The lack of the jaw margins and the specialized tooth plates 

 of the fossil forms supports such a proposition. 



The Dipnoi are peculiar among the groups so far exam- 

 ined in that they appear to lack some of the common skeletal 

 inheritance of the choanates and appear to have lost much 

 of what they had acquired during their evolution. It has 

 been suggested that this is not a case of loss but, to the con- 

 trary, that in this group the full pattern was never present. 

 Such a view must be based on the assumption that the 

 tetrapod pattern actually arose among the choanate fishes 

 and that one branch of this ancestral group, the Dipnoi, di- 

 verged from the stem before the development of that pattern. 

 Any conclusion regarding this possibility must await a con- 

 sideration of the skeletons of other fishes. 



ORIGIN OF THE AMPHIBIAN 

 OR TETRAPOD 



Of the two groups, crossopterygian or dipnoan, which is 

 closer to the amphibian? .Save-Soderbergh has pointed out 

 the unlikelihood of the origin of amphibians directly from 

 crossopterygians. This view can be based on the lack of 

 agreement in the dermal bones of the snout, specifically the 

 presence of a frontal and nasal as opposed to a series of nasals, 

 and also on the autostylic palatoquadrate of the amphibian 

 and the possible loss of the hyomandibula. 



Holmgren has assumed a dipnoan origin for the urodeles 

 in spite of the obvious specializations of the Dipnoi when 

 first encountered in the fossil record. Even though Jarvik's 

 analysis of the nasal openings of dipnoans is discounted, the 

 loss of maxilla and premaxiUa and the form of the pterygoid 

 and prearticular dental plates remain as salient features. 

 Offering some support to the dipnoan origin are the auto- 

 stylic jaw suspension, reduction of the hyomandibula, and 

 connection of the pharyngosuprahyal to the otic capsule. 

 Also lending credence to this view is WestoU's use of the cra- 

 nial pattern of the early dipnoan as suggestive of the ances- 

 tral type for the choanates. 



To complicate the discussion of the origin of amphibians, 

 Save-Soderbergh, Holmgren, and Jarvik have asserted that 

 the Amphibia arose from two choanate-fish sources. Save- 

 Soderbergh and Holmgren derive the Urodela from a source 

 near the Dipnoi and the other amphibians indirectly from 

 the osteolepid crossopterygians. Jarvik derives the urodeles 

 from a porolepiform and the anurans and other tetrapods 

 from an osteolepiform source. The arguments supporting 

 these viewpoints are long and detailed. Neither assumption 

 is particularly supported by the foregoing accounts. 



The origin of the amphibian is much more difficult to 

 document than that of the reptile, because, we have been 

 moving back further in the time scale with each of the 

 groups considered and are now near the beginning of the 

 fossil record. Although there is little doubt that a relation- 

 ship exists between the early "choanate fishes'" and the 

 "choanate" amphibians, assessing this relationship is impos- 

 sible because of the lack of fossil material. The Amphibia, 

 when first encountered in the fossil record, had already 

 undergone a great deal of radiation and were perhaps con- 

 temporaneous with reptiles. 



104 • HEAD SKELETON OF OTHER TETRAPODS AND CHOANATES 



