iAConiN.i:. 



reasons however, it is convenient to again split up the first division according as the lateral line is 

 ventral or ventral-niediolateral (cf. the table of analysis p. n — 12). 



We may now refer to some of the characters which have special importance in distinguishing 

 the species within the greater groups based on the course of the lateral line. 



Scales. Of the present species two are perfectly devoid of scales, namely Lycodes agnostus 

 Jensen (Tab. VI, fig. 1) and L. platyrhiims Jensen (Tab. VI, fig. 2). The well-known ichthyologist P. 

 Hleeker has laid such great weight on the absence of scales that lie has formed a special genus 

 Lycodaiepis 1 ), which only differs from Lycodes in this one character, and the later American naturalists 

 have followed him. It seems to me that Lycodalepis is an unnatural genus. If the two scaleless 

 species mentioned are removed from the genus Lycodes, they are then separated from species to which 

 in other respects they are closely allied. L.agnostus has its true place amongst the species of Lycodes 

 with mediolateral lateral line, not only on account of the situation of the lateral line, but also on 

 account of the relatively short tail (the head and trunk together are 46 — 52"/,, of the total length) and 

 the number of rays in the unpaired fins (I). go — 93, A. 70- 72). L. platyrhinus on the other hand, 

 belongs rightly to the Lycodes with double lateral line, partly because of the veutral-mediolateral 

 lateral lines, partly because of the long tail (head and trunk together are 37% of the total length » 

 and the number of rays in the unpaired fins (D. 0,9, A. 82). One must be content therefore, to regard 

 the absence of scales as a good specific character and not ascribe to it any generic importance. 



( >f the remaining species of Lycodes dealt with in this treatise, there is one which in its 

 slight development of the scaly covering, is a transition form to the naked species, namely /.. seminudus 

 Reinh. As the name denotes, it is only half covered with scales, on the tail namely 2 ), and moreover 

 the scaly area ends in front in the shape of a wedge leaving a naked part dorsallv and veutrally 

 (Tab. IX & Tab. X, fig. i). Some variation appears in this species however, as the scaly wedge some- 

 times sends a portion forward on to the trunk; but as a rule the naked abdominal region is one' of 

 the characters which aids to a ready determination of L. seminudus. 



The naked L, agnostus and the half naked J., seminudus belong, as mentioned, to the group 

 with mediolateral lateral line. The remaining species of this group, L. rossi Malmgr., L. lutkenii Coll. 

 and L. reticulars Reinh. have the scaly covering developed almost to the same extent, and it extends tor- 

 wards on the sides of the trunk to a point which lies under or a little in front of the beginning 

 of the dorsal fin; but the front part of the back and the belly are always naked, and the fins ar« 

 likewise devoid of scales. 



Most of the species of the groups with ventral or veutral-mediolateral lateral lines are remark- 

 able for the stronger development of the scab' covering. In adult individuals the scales covei the 

 whole of the body (head excluded) and extend on to the unpaired fins. Such is the case in L.vaklii 

 Reinh., L. frigidus Coll. , L. atlanticus Jensen, L. esmarkii Coll. and L. eudipleurostictus Jensen. Of 

 these species L./rigidus is easilv recognisable by its extremely small scales (Tab. V, fig. 1 a). L.palhdus 



'I Versl. K. A. W. 2r Rks. VIII, 1S74, p. 369 (mucosus). 



) Jordan & Evermann form a subgenus Lycias under Lycodes, characterised by the tail onh being scaled; into 

 tins subgenus are brought L. seminudus Reinh. and the imperfectly described of I-Croyer, 



rediscovered (I-'isht-s of North and Middle America, III, 1898, \>. 2463). 



