EKL-FI^IIKS. 



1017 



Tlie tliinl iiiterpri'tatiuii (3) — accordiiiy' t(i wiiicli 

 the margin ol' tlif upper jaw consists laterally of the inter- 

 iiiaxillaries and anteriorly of the vomer (there coaleseent 

 with tlic etlinioid) — was first proposed l)y C'uvikh", 

 and lias long been the most genei'ally accepted. It j)re- 

 supposes that true maxillaries and palatines are wanting. 



The history of tiie Eels' dexelopnient has not \'et 

 pronounced judgment on these three oj)inions; and a 

 fourtli explanation migiit reasonably be suggested, name- 

 Iv that the palatines and maxillaries were originally 

 independent, but have coalesced, as Meckel and, after 

 hini, 1'etei!s assumed of the relation between the inter- 

 inaxillaries and the etlimoid and vomer. Fi'oni the 

 evolutional instory of tlie higher vertelirates we know'' 

 that the maxillaries as well as the ])alatine roof are 

 primordial excrescences of the mandibular arch, and 

 tiiat tile iiicii)ieiit palatine roof ma^' appt^ar as a se- 



\\.r 



Fisf. 271. Hyoid arch of Anguilla vulyaris. X '' .,. 

 A: liyiiiil arch wilh Iminchiostegal rays; B: linsjunl Ixmc; (': tlic 



seen from below. 

 /, lingiinl; iih, iiroliyoid; bbr, first copula (hasi-braiichi.il); cell, 

 tohyoid: c/i/f, cpihyoid; R. br., brancliio.'itefral rnys. 



condary growth in an inward direction, originating from 

 the maxillary arch. \Miether an equally intimate rela- 

 tion between these parts has possibly arisen during their 

 reduction in the Eels, is a question which onl}' the his- 

 tory of evolution can solve. We accordingly consider 

 it still doubtful whether intermaxillaries and maxillaries 

 are wanting in the Eels or have coalesced with more 

 internal btnies, and in the present state of the (juestion 

 the describer is fully entitled to speak of maxillary 

 and vomerine teeth. 



In conjunction witli the great development of the 

 branchial cavity the Eels have very broad branchiostegal 



" Man. Mils. d'Hist. Nat., tome I, p. 118. 



'' Cf. Smitt, L'r de hogre djureni' iiti'ecklh>i].'!histo)-ia, p. 167. 



mcmbrimes and long branchiostegal i'a\"s (A', hr.). iSiit 

 the foi'mer arc united ihrougiiout by far the greater 

 part of their margins to tlie skin and joined to each 

 otiiei', thus leaving oni\' small, slit-like gill-openings; 

 and all the latter, excejit the uppermost (hindmost) ones, 

 are slender, filanumtous, and coiled backwards, upwards, 

 forwards, and rinidl\- dnwiiwards, so as to include within 

 their cur\c both the operculum and the suboi)erculum. 

 In the Conger (fig. 2X2, p. 103G) the last two bran- 

 chiostegal rays of each membrane are expanded at the 

 top; in the common Eel (fig. 271) the expansion is con- 

 fined to the last ray, lint is so great that the ray is 

 sometimes e(|iial in breadth to the comijaratively nar- 

 row, obliquely semi-elliptical operculum (fig. 270, op). 

 The sabre-shaped — in the Conger (fig. 282) elongated 

 triangular, posteriori)' pointed, but curved — suboper- 

 culum (sojj) follows the curved inferior margin of the 

 oyjerculum, and is weaker iioth in the common Eel and 

 the Conger than the triangular, anteriorly pointed inter- 

 oi)erculum (iojj), which is even larger than the thin, 

 crescent-sha])ed preoperculum {x>op). 



The Eels are exceedingly voracious; and their di- 

 gestive canal is accordingly rather short and simple. 

 Our common Eel raaj' serve as an example. The main 

 abdominal cavity of this species varies considerably in 

 length, between about 26 and 30 % of the length of 

 the body; but in the caudal part it has on each side 

 a prolongation, of which we shall give a fuller descrip- 

 tion when treating of the sexual organs. The thin- 

 walled and rather wide (esophagus passes into the sto- 

 mach ivithout a break, except that the longitudinal 

 folds (ridges) in whicli the mucous membrane lies on 

 the inside become thicker and higher in the stomach, 

 while the wall of the stomach is also thicker than that 

 of the oesophagus. Straight backwards the stomach is 

 prolongated in the form of a cone; but from the right 

 and lower side of its anterior part it sends out a pyloric 

 portion, with e(]ually thick walls, which runs forward 

 to the hind margin of the liver, where it bends abruptl)' 

 in a narrow crook, so that the beginning of the intestine, 

 which runs backwards, lies close to its under surface. 

 There are no cajcal diverticula. .\t the middle of the 

 crook is a funnel-like valve (an annular fold on the in- 

 side), marking off the pvlorus from the intestine; and 

 on the inside of the latter the mucous uiemlirane is 

 di.sposed in a honeycombed network of dee]), confluent 



