the present. The position of the nasohypophyseal opening 

 on the top of the head in the osteostracan and anaspid sug- 

 gests relationship with the lamprey. The position of this 

 opening at the tip of the snout in the myxinid, along with 

 the general separation of cranial and branchial skeletons 

 suggest as close a relationship between the hagfish and 

 lamprey as with the heterostracan. The many differences in 

 structure of the hagfish as compared with the lamprey sug- 

 gest that the two living lines arose at the time of the pri- 

 mary radiation of the agnaths, or shortly thereafter. 



The relationship of agnath to gnathostome is hard to de- 

 cide. Johann Miiller (1839), the first student of the myxinid 

 type, decided that the "Bauplan" of the marsipobranchs 

 (pouch-gilled fishes) was different from that of the gnatho- 

 stome. In the pre-Darwinian period, such a decision could 

 be made without serious difficulties, but this explanation 

 was unacceptable to the evolutionists. Cole (1905) was able 

 to account for the lack of jaws and the lack of similarity to 

 the gnathostome "type" by identifying the head skeleton of 

 the myxinid as a functional neomorph, that is, a skeleton 

 developed in this type to meet its own needs. 



The theory of Gegenbaur, that the gnathostome jaws 

 were developed from anterior gill arches, saw in the agnath 

 a stage preceding that of the gnathostome. There is little 

 evidence of visceral arches being involved in the mouth 

 area of the agnath — the evidence is more apparent in the 

 gnathostome. In spite of the difficulties, many comparisons 

 have been drawn between the mouth parts of these two 

 groups. If the blastematic condensations observed by Holm- 

 gren, a recent contributor to this problem, are equated to 



palatoquadrate, Meckel's cartilage, and hyoid arch, then 

 there is evidence of gnathostome jaws in this agnath group. 

 A correlation of parts with jaws, whether following Holm- 

 gren's views or those of Ayers (1921, 1931), leads to the con- 

 clusion that the gnathostome structure preceded the agnath. 

 Deriving gnathostomes from agnaths, or the reverse, neces- 

 sitates the formation of many tenuous homologies accom- 

 panied by lengthy and often obscuring explanations. The 

 following summary of views regarding the visceral skeleton 

 indicate this. As early as 1832, Rathke suggested that the 

 branchial arches of the lamprey were homologous with the 

 extrabranchials of sharks. Balfour in 1881 expressed the 

 same opinion. However, Huxley (1875-76), Parker (1883), 

 and Howes (1891-92) considered that the visceral arches in 

 cyclostomes and gnathostomes were homologous. Dohm 

 (1884) tried to show that the superficial position of the 

 branchial arch of Petromyzon, as compared to that of sharks, 

 was to be regarded as a secondary phenomenon — merely a 

 shift in relationships— and that, consequently, no objection 

 could be raised to homologization. Against Dohrn's opinion, 

 Goette (1901) stated that the branchial arch of the lamprey 

 was fundamentally different from that of the gnathostome 

 because of its superficial position. Sewertzoff (1916-17) en- 

 deavored to trace the origin of gnathostomes and cyclo- 

 stomes back to a common ancestor from which both styles 

 could have evolved — evolution from a common ancestor 

 demanded homology in his opinion. Neumayer (1938) 

 agreed with Sewertzoff, while Allis (1923-24) tried to 

 homologize cranial elements of cyclostomes with extra 

 visceral elements in gnathostomes. Damas (1944) stated 



rostral 



orbital plates 



mouth 



oral plates 



preorogoniales 



orogoniole^ 



dorsal plate 



ventral plati 



preana 



cornual plate 



branchial plate. 



dorsal spines 



ronchiol opening 



anal spines 



Figure 5-33. Anterior part of body of Orepanospis gemoendensis as seen from above, left, and 

 below, right. (Mainly after Obrutschev, 1943) 



AGNATH FISHES • 139 



