Genera/ observations The head skeleton of the antiarch 

 only superficially resembles that of the euarthrodires. In 

 view of the radical differences in design of the several sub- 

 divisions of arthrodires, it is quite possible that the acan- 

 thodians and sharks could be included in the same group. 

 Such inclusion would be relatively meaningless in terms of 

 comparisons with the osteichthian fishes. Even if the acan- 

 thodians and sharks are not included, the Arthrodira as a 

 group is scarcely definable. 



Comparisons of the shark type (with or without Hydro- 

 lagus) with the acanthodian or arthrodire is extremely diffi- 

 cult. Whereas the shark is characterized by lack of bone, 

 the arthrodire is generally a heavily armored type — with a 

 roof pattern having some resemblance to the osteichthyian. 

 Detailed comparisons are not yet possible because of lack 

 of information on these fossil groups. 



Interrelationships at the base of the 

 gnathostome Tishes 



The term teleostome has been used to identify actinopte- 

 rygians and choanates having "perfect" or complete jaws, 

 i.e. there is a premaxillary and maxillary margin to the 

 upper jaw and a dentary margin for the lower jaw. These 

 bones bear the main teeth or tooth rows. In contrast, the 

 plagiostome mouth, meaning oblique or transverse, lacks 

 this margin. As far as known, only the palatoquadrate car- 

 tilage is involved in the plagiostome upper jaw, and Meckel's 

 cartilage in the lower jaw. The acanthodians and arthrodires 

 may lie somewhere in between these two extremes, for they 

 have vomer and palatopterygoid teeth and dermal jaw 

 margins. 



In terms of the branchial skeleton, the teleostomes differ 

 from the plagiostomes. The pharyngobranchial extends 

 forward and medially in the teleostomes and posteriorly and 

 medially in the plagiostomes. Again they differ in the struc- 

 ture of the "hyomandibula" and the presence of a symplec- 

 tic, or interhyal. 



The number of branchial arches is generally five in both 

 of these divisions; exceptions include several of the sharks 

 which have six or seven. This increased number is probably 

 due to the addition of slits rather than retention of a primi- 

 tive number. 



The number of segments in the branchial arches appears 

 to be a constant. The hyoid arch of gnathostome fishes is as- 

 sumed to be only a modified branchial arch. The apheto- 

 hyoidean is presumed to be more primitive than those types 

 m which the epihyal is modified by participation in support 

 of the lower jaw. 



The basic branchial arch plan has been extended to the 

 mandibular and premandibular arches by Jarvik (1954) in 

 support of Gegenbaur's hypothesis. In this scheme, the 

 palatoquadrate is seen to be a compoimd structure with 

 autopalatine and quadrate subdivisions. The quadrate part 

 is the epimandibula, and its ascending process (epipterygoid 



or metapterygoid) is the supramandibula. The otic process, 

 comparable to the laterohyal, can be termed the latero- 

 mandibula. The trabecula is the pharyngo element. The 

 premandibular arch has as its epi element the autopalatine, 

 or processus pterygoideus. This part of the palatoquadrate 

 lies anterolateral to the prespiracular groove and thus ante- 

 rolateral to the diverticulum in the roof of the pharynx, 

 which is interpreted as a vestige of the premandibular gill 

 pouch. The orbitonasal lamina of the endocranium is the 

 supra element, while the intermediate mesenchyme below 

 the ethmoid region, which gives rise to the vomer, represents 

 the pharyngopremandibula. 



Supporting Jarvik's associations are the findings that, 

 while the parachordals and polar cartilages are entomeso- 

 dermal (i.e. of sclerotome or dermomyotome origin — see 

 Chapter 7), the trabeculae are of neural crest origin or ecto- 

 mesodermal (thought to be both ecto- and entomesodermal 

 in sharks), like the various jaw and branchial arches. 



Basically, the plagiostome agrees with the teleostome in 

 most details of the endocranium and in the general course 

 of development. The plagiostome differs in that the para- 

 chordals lie close to the notochord throughout their length; 

 the notochord projects through the acrochordal or dorsum 

 sellae; there is a strong angle between the parachordals and 

 trabeculae; the palatoquadrates lack the autopalatine com- 

 ponents and join below the trabeculae to form the upper 

 jaw; there are circumorbital cartilages (or mesenchyme) and 

 separate nasal cartilages (these may correspond to the para- 

 nasals of teleostomes); and a lateral commissure is missing 

 or vestigial. These and other features indicate the basic 

 dichotomy of the vertebrate stem into plagiostome and 

 teleostome lines. However, in terms of ossification and prob- 

 ably in terms of dermal jaws and many of the other features 

 considered above, the plagiostome and teleostome are prob- 

 ably only the extremes of a series represented in part by the 

 acanthodians and arthrodires. 



AGNATH FISHES 



The jawless fishes, or agnaths, are presumed to be the most 

 primitive of the vertebrates, preceding the gnathostomes in 

 time and indeed giving rise to that group. The living 

 agnaths, the cyclostomes, are considered relict species, highly 

 specialized in their way of life — a way which has protected 

 them from competition with gnathostomes. 



Cyclostomes 



There are two kinds of cyclostomes, the lampreys (repre- 

 sented by Petrnmyzon) and the hagfishes (Myxme and Eptalre- 

 tus). Both have cartilaginous skeletons. They lack any trace 

 of bone, and even the teeth are horny. These two types are 

 so different that both should be considered. A description of 

 these is difficult because the structural landmarks of the 

 gnathostome are almost completely lacking or altered in 

 interrelationships. 



128 • THE HEAD SKELETON OF FISHES 



