pseudobranch (mandibular hemibranch) or hyoidean heini- 

 branch. The opercular structure, in terms of bones, is dis- 

 tinctive, as has already been pointed out. 



The larva has a pair of large external gills which grow 

 back from the hyoid arches— and receive their blood supply 

 from those arches. This gill has dorsal and ventral rows of 

 hlaments. It appears before the branchial openings and per- 

 sists relatively late in development, even after the formation 

 of the operculum. 



In terms of its respiratory apparatus, Polypterus appears 

 to be quite distinct from the actinopterygians. Lack of the 

 pseudobranch and hyoidean hemibranch could be explained 

 as due to the respiratory function of the air sacs. However, 

 the bilobed ventral "lungs" are also unique among actinop- 

 terygians, even if one discounts them as representing the 

 primitive type from which the actinopterygians diverged. 

 The larval gill of Polypterus is unmatched in any other type 

 offish (Figure 7-21). Again, it could be considered a primi- 

 tive type of actinopterygian external gill which has been 

 lost in all other members of this group. 



Chondrichthyes Sharks have a spiracle with a mandibular 

 pseudobranch, a hyoidean hemibranch and four, five, or six 

 holobranchs (Figure 11-22). The posterior branchial open- 

 ing lacks a posterior hemibranch. Each of the branchial 

 clefts open separately to the outside. The medial septum ex- 

 tends beyond the gill filaments forming a flap covering the 

 branchial slit behind. Muscle fibers extend down through 

 this septum. A single row of gill rays supports the septum as 

 compared with the double rows of gill rays supporting the 

 two rows of filaments of the actinopterygian holobranch. 

 Polypterus is like the actinopterygian in this feature, while 

 the choanate lacks gill rays. This lack of gill rays is probably 

 related to the reduction of ossification and development of 

 the branchial system in the choanate group. 



The branchial arches of the shark have one afferent ar- 

 terial channel and bilateral efferent channels, as does the 

 dipnoan Neocemtodus. Aapenser is intermediate with respect 

 to the teleost, holostean, and Polypterus, in which there is one 

 afferent and one efferent vessel. 



The holocephalan is operculate. The opercula are joined 

 below and with the isthmus. There are four branchial open- 

 ings (Figure 1 1-24). The spiracular cleft is completely closed, 

 although a spiracle is present in the young. There is no 

 evidence of a pseudobranch in the spiracular pouch. The 

 hyoid bears a posterior hemibranch fused to the inner sur- 

 face of the operculum. The first, second, and third arches 

 have holobranchs and the fourth a hemibranch. The gill 

 filaments reach outward just beyond the septum. Each arch 

 has only one efferent vessel and there are valves along its 

 anterior and posterior margins. These valves act to close the 

 branchial openings when the mouth is opened. The opercu- 

 lum serves this function in the actinopterygian. 



The holocephalan thus differs from the shark or actinop- 

 terygian in its branchial organization. 



Fossi/ gnathosfome fishes Little is known about the bran- 

 chial skeleton of the arthrodires and therefore of their gills. 

 They are usually described as aphetohyoideans but, using 

 the holocephalan as an example, this does not mean that 

 there was a full mandibular slit or even a spiracle. The 

 head shield extended back over the branchial region but 

 did not form a movable (functional) operculum. 



The acanthodians had three or four (rarely five) branchial 

 arches and they too had a separate (not jaw supporting) 

 and complete hyoid arch (the aphetohyoidean condition). 

 There was a mandibular operculum supported by rays. The 

 hyoid arch, and the more anterior branchial arches, may 

 have had small dorsal opercular flaps. 



Cyc/ostomes The lamprey has seven pairs of branchial 

 pouches, beginning with the second branchial pouch — that 

 one lying in front of the second branchial arch. Each of these 

 pouches is lined with gill filaments and served by afferent 

 and efferent arteries lying between the pouches. The phar- 

 ynx of the adult lamprey is divided into a dorsal esophageal 

 and a ventral respiratory passage (Figure 9-26). A narrow 

 duct leads from the respiratory pharynx into the more or 

 less spherical pouch. The pouch opens to the exterior through 

 a small, round opening guarded by valves. 



In the larval lamprey the pharynx is undivided and opens 

 widely into or between the branchial arches. The pouches 

 open to the exterior through small round ports. During met- 

 amorphosis the pharynx is divided and the constricted in- 

 ternal openings of the branchial pouches are developed. 



The myxinoid is quite different from the lamprey, shar- 

 ing only the pouch-like nature of the branchial clefts. (These 

 fishes have been called the marsipobranchs on this account.) 

 The number of pouches varies in this group; there are six 

 pairs in Myxme and as many as thirteen to fifteen pairs in 

 Eptatretus. The first branchial pouch never appears, the sec- 

 ond develops, but degenerates as does another more pos- 

 terior pair in the course of development. It is not known 

 which pair of pouches becomes the most anterior of the 

 aduh, but the branchial apparatus is displaced far poste- 

 riorly. The myxinoid is peculiar in that a pharyngocutaneous 

 duct leads from the pharynx to the exterior on the left side 

 behind the last pair of gill pouches. In Eptatretus each gill 

 pouch opens separately to the exterior, while in Myxine the 

 excurrent channels join and open together behind the level 

 of the gill pouches. The pharyngocutaneous duct opens widi 

 this left channel in Myxine (Figure 9-28). 



The myxinoid differs from the lamprey in that the phar- 

 ynx is undivided (Figure 9-27). The vascular supply of the 

 pouches is also peculiar in that the afferent and efferent 

 arteries serve a single pouch only— they are not spilt be- 

 tween two pouches as in the lamprey (Figure 1 1-26). 



The functioning of the gill pouches of the lamprey and 

 myxinoid is complex. When the mouth is attached or 

 when the head is buried in a victim, as in the case oi Myx- 

 ine, some point of entry other than the mouth is necessary. 



THE RESPIRATORY SYSTEM • 289 



