fishes did not have a complete mandibular cleft (the canal 

 could not have been displaced forward on the cheek if such 

 a cleft were present). In some acanthodians a dorsal hyoid 

 section extends from the spiracle to the temporal canal. 



The supraorbital canal of the acanthodian may be con- 

 nected with the lateral line or may be separate. Occipital 

 and posterior pit-line canals are indicated in this group. 

 The snout situation is not known. 



In the arthrodires the canal system is indicated by grooves 

 in the dermal bones (Figure 5-22, 13-25 A); this system is 

 peculiar because of the radiating nature of the supraorbital, 

 infraorbital, and lateral lines. This is also observed in the shark 

 Laemargus. Unlike any other gnathstomes, there are also 

 middle and posterior pit lines in this radiation. The posterior 

 pit line is a canal in this group; the occipital canal, passing 

 behind the endolymphatic pore, is lacking or poorly devel- 

 oped. The temporal line loops down toward the cheek and 

 the preopercular canal is indicated only dorsally. A jugal 

 connection between mandibular and hyoid arch lines is per- 

 haps present in Homosteus. Like the shark the supraorbital 

 canal extends down behind the nasal capsule, but it does 

 not join the temporal line posteriorly. It is unlikely that the 

 infraorbital canal passed below the nasal capsule. 



Act'inopterygian fishes The sensory-canal system of the 

 actinopterygian fishes is difficult to compare with the hypo- 

 thetical pattern or the patterns seen in other groups (Figure 

 13-30). As in the chondrichthyes, most acanthodians, and 

 arthrodires, there is no evidence of a reticular canal system in 

 any of the living or extinct forms. The supraorbital canal 

 may or may not connect with the temporal division of the 



lateral-line canal. The infraorbital canal passes forward be- 

 low the nasal capsule; anterior connection between the 

 supraorbital and infraorbital canals is lacking with the ex- 

 ception of Polypterus, The hyoid arch canal attaches to the 

 temporal canal behind the spiracle or ends on a line passing 

 behind the spiracle. The mandibular arch line is poorly 

 developed; it is suggested only by horizontal and vertical 

 pit lines on the cheek. There is an occipital canal, and ante- 

 rior, middle, and posterior pit lines on the top of the head. 

 The anterior pit line is a continuation of the supraorbital 

 canal behind the point of connection with the temporal 

 canal. This suggests that it is a part of the dorsal line as 

 seen in the agnath fish or the basic model. 



Not all actinopterygians agree in detail. In the palaeonis- 

 coids, Pteronisculus or Boreosomus as examples, the supraor- 

 bital canal does not join the temporal canal. This state is 

 observed in all lines of palaeoniscoid fishes. The living 

 chondrosteans, holosteans, and teleosts in contrast have 

 these lines joined. Although the preopercular line invariably 

 extends up behind the spiracle, Polyodon is the exception. In 

 this chondrostean it connects with the infraorbital line be- 

 low the spiracle. In the haplolepid palaeoniscoids, dorsal 

 and ventral body lines as well as a main lateral line are ob- 

 served. 



In the actinopterygian the sensory lines are usually em- 

 bedded in the dermal bones with pores to the surface. The 

 dermal bones may be reduced to tubes enclosing the canal. 

 In some the canals lie superficial to the bone or are lost. 



The development of this system in Anna has been re- 

 ported (Figure 13-31). Here again are seen supraorbital, 

 infraorbital, joined mandibular and hyoid, and temporal 

 placodes. 



posterior pit line 

 middle pit line \ ^occipitol canal 



anterior pit line 



supraorbital canal 



anterior nasal tube 



dorsal sensory line organs 



lateral line 



mandibular division' 



gular pit line' 



Figure 13-30. Sensory lines, organs, and innervation of organs in Amia. (After Allis, 1888-89) 



SENSORY ORGANS 



415 



