180 MK. F. J. COLE OX THE STRUCTURE AND MORPHOLOGY OF 



There are supra-temporal and supra-orbital commissures, and opposite tlie latter on each 

 side is a large l)lind ciecum, corresponding' exactly to the smaller one found by me in 

 vounff Grnlus vireiis. As three of these cjieca are also found on the infra-orbital canal 

 (two posterior and one ventral to eye) in positions occupied in other forms by dermal 

 tuhules, they doubtless correspond, as I surmised above in describing the one found in 

 G. vireiis, to modified dermal tulmles. The sensory canals of Lota are, further, largely 

 ampulliform. In Salmo, "VV. K. Parker (1S73, 152), who regarded the sensory canals as 

 fflandular organs, describes the hyomandibular canal as communicating with the lateral 

 or body canal (p. 99), the Salmon in this respect differing from Gadiis. 



Prom the description given by Emery (1880, 66) it is (\isy to derive the lateral 

 canals of Fierasfer ecus from the Gadtis type. There are, however, some differences, 

 of Avhich the most important are : — (1) there is an anterior infra- and supra-orbital 

 anastomosis, which is an Elasmobranchian rather than a Teleostean character — the 

 anastomosis in Amia is between the two infra-orbital canals; (2) the two supra-temporals 

 anastomose across the back to form a supra-temporal commissure, Avhich is situated 

 immediately opposite the operculo-mandibular canal and leaves tlie body canal on the 

 squamosal : the innervation, howeA^er, is precisely as in Gadiis ; (3) there is a supra-orbital 

 commissure, but the median unpaired tubule is represented by two posteriorly directed 

 blind sacs ; (4) the operculo-mandibular canal does not end blindly behind, but opens 

 into the main canal of the head ; (5) the dermal tubrdes are much reduced in number, 

 and Avhere present are large, blind, and ampulliform as described by Hyrtl in Lota. 

 Fierasfer dentalus agrees with F. aciis, hut represents a still simpler condition. To put 

 it briefly, Fierasfer represents a simplified edition of the Cod type. 



Sappey (1880, 175) figures (pi. xi. figs. 1 & 2) and describes (p. 43) a very curious 

 condition of the sensory canals in the " congre commun " [Conger vulgaris). The canals 

 are reduced to the two lateral or body canals, and a very short operculo-mandibular 

 canal, Avhich communicates with the body canal on each side and terminates opposite 

 The angle of the jaw in a large ampulla. In Cottus, according to Bodenstcin (1882, 24), 

 the geography of the sensory canals is much the same as it is in the Cod. There is a 

 supra-orbital commissure giving off a median dermal tubule as in Gadns, but the two 

 •supra-temporal canals also form a commissure. The operculo-mandibular canal does not 

 anastomose with the body canal. There are fewer dermal tubules, and conseqviently 

 there must be fewer sense organs. In Amiurus Eamsay Wright (1884, 227) describes 

 an occipital commissure, but failed to find one between the supra-orbital canals. The 

 operculo-mandibular canal is separate from the body canal. Fritseh states (1887, 75) 

 that the lateralis lateral line nerve consists of two (upper and lower) rami as in Gadrn, 

 which he calls respectively the R. superticialis lateralis vagi and the R. profundus 

 lateralis vagi. 



An examination of Allis's first Amia paper (1889, 4) discloses many points of resemblance 

 between that fish and Gadtis. We have seen that in the latter fish the sensory canals 

 have a ligamentous support only in two places. In Amia, Allis says (p. 469) : — " In 

 their passage from one bone to another, where the bones are not naturally connected, 

 the canals lie in a dense connective tissue, which forms the deeper part of the cutis. 



