40 



ME. G. SIM OK" THE OCCUERENCE ETC. OF 



of dentitioD, Centronotus tas a single row of conical teeth in both 

 jaws, becoming shorter as they approach the angle of the mouth ; 

 sometimes there exist a few irregularly set teeth in the upper jaw 

 in addition to the regular row ; several strong teeth on the 

 vomer ; four patches, two on each side, of irregularly set upper 

 pharyngeal teeth, and two long patches below, also set irregu- 

 larly. On the other hand, Lumpenus is furnished in the upper 

 jaw with numerous irregularly set teeth, the outer ones being 

 conical, the others sharp-pointed, and all decreasing in length as 

 they approach the angle of the mouth. The lower jaw is set with 

 one, sometimes two rows of long incurved, rather sharp teeth, 

 not of uniform length, and all becoming shorter laterally ; none 

 on the vomer or tongue. There are two irregularly set patches 

 of upper pharyngeal teeth, each point of which is finely tinged 

 with yellow ; and four lower pharyngeals forming two sides of a 

 triangle set confusedly, and without the yellow colour of the 

 upper ones. In the number of vertebrae, Centronotus ouirnxmbev^ 

 Lumpenus by 3, the latter having 80 and the former 83. 



But the greatest departure which Centronotus makes from 

 Lumpenus is in the form of the stomach and intestines. As will 

 be seen by the accompanying figures, the intestines of Centronotus 

 in some cases show little change in the stomach from the intestines 

 proper ; that is, there is no very distinct line of demarcation 

 between them ; presenting more or less the form of a simple 

 tube and are devoid of caeca. I have said in some cases. This 

 will be best understood by referring to figs. 1 and 2, both of 

 which represent the intestines of Centronotus, both drawn from 

 specimens caught by myself and while the intestines lay in situ. 

 This is a change in the internal arrangement of a species I have 

 never before observed carried to such a length, although the indi- 

 viduals of no species have the intestines formed or laid exactly 

 alike. The result of my investigation goes to show that fig. 1, 

 or something much like it, is the most usual arrangement to be 

 met with in Centronotus gun7iellus. On the other hand, Lum- 

 penus has what may be called a properly developed stomach, 

 marked ofi" from the intestines by its two caecal appendages, and 

 the intestinal track forms several convolutions in its course. 



With reference to the caeca in Lumpenus^ I find they differ 

 much in length in diff'erent individuals, even although these indi- 

 viduals are of themselves the same size. In some examples 



