NATURAL SCIENCES OF PHILADELPHIA. 517 



In other families, such as the Chromoids, Pomacentroids and Labroids, the 

 lateral line is also generally interrupted, or it is sometimes obsolete behind, 

 and sometimes the superior or anterior portion is deflected and join3 the infe- 

 rior and posterior one, thus forming a continuous line, but not pursuing an 

 uninterrupted course. No instance is apparently known where the superior 

 part is entirely developed at the expense of the inferior. 



In other respects, the likeness of Notothenia and ELeginus is very apparent, 

 and was commented on by Sir John Richardson; yet he, with too implicit re- 

 liance on Cuvier, retained the latter in the family of Scitenoids, while he 

 referred the former to that of Gubioids on account of the flexibility of the 

 dorsal spines of the typical species, and the resemblance to Eleotris. All of 

 the essential characters of Notothenia and Eleginus appear, however, to be 

 shared, the differences being indicative of much less than family rank. The 

 general physiognomy, the essential structure of the fins and the anatomical 

 features, as far as known, are similar. The system of muciferous pores is also 

 entirely homologous; it is most app irent in ELeginus, and may be illustrated 

 here. 



The lateral line is' continued forwards by pores nearly along the oculo- 

 scapular groove, beneath the eyes and near the margin of the snout; again, 

 along the margins of the preopercula and beneath the dentaiy bones. At the 

 nape are generally three pores between those of the oculo-scapular grooves. 

 Five others are arranged in a more or less regular quincunx in and behind the 

 interorbital area: between each one, forming the anterior angle of the quin- 

 cunx and the tubular nostril, is generally another. Supernumerary ones are 

 frequently interposed between the preceding; more rarely all on the crown 

 and forehead are rudimentary or obsolete, as in the genus Macronotothen. 



It remains to indicate the distinctions between the family and others. From 

 the Sciaenoids it is distinguished by the non-development of a cavernous skull, 

 the increased number of vertebra;, the structure of the fins and the extent of 

 the anal, the position of the ventral fins, the number of branchiostegal rays, 

 the course of the lateral line, and the absence of scales on the fins. The physi- 

 ognomy is also quite different. 



It differs from that of the Gobioids by the form of the stomach and the pre- 

 sence of pancreatic cajca, the structure and extent of the vertical fins, the 

 relation of the ventral ones, the presence of a lateral line and the structure of 

 the scales. 



It is more nearly allied to the Latiloidae, the Harpagiferoidas and the Chae- 

 nichthyoidae, but its physiognomy is unlike that of any of them. From the 

 first, it is also distinguished by the structure of the fins, the course of the 

 lateral line and perhaps the development of the muciferous pores. From the 

 Harpagiferoids, by the scaly body and the unarmed head. From the Chasnich- 

 thyoidae, by the scaly body, as well as the form and structure of the head and 

 the normal development of the snout. 



It might, perhaps, be natural to unite the two last named form3 in the same 

 family with the present, but the different aspect of the respective groups joined 

 to the positive characters distinguishing them, scarcely appear to warrant 

 such a combination. The Harpagiferoids are in some respects the most closely 

 allied. The Notothenioids seem to hold a position in the Southern seas analo- 

 gous to that of the codfishes of the Northern ones. 



Subfamily NOTOTHENIIN.E (Gthr.) Gill. 



Synonymy. 



Nototheniina, part. (Notothenia) Gilnther, Annals and Magazine of Natural 

 History. 



Notolhenioidce corpore regulariter attenuata, linea laterali postice interrupta. 



Body robust and anteriorly subcylindrical. Lateral line interrupted behind. 

 1861.] 



