EEPORT ON THE DEEP-SEA FISHES. 123 



in firmness, those of the infraorbital ring cover more or less the side of the head, extend 

 backwards to the angle of the prseoperculum, and push the latter backwards. The snout 

 becomes the receptacle of large or even enormously enlarged cavities, supported by thin 

 osseous ridges, and projects more or less beyond the mouth, which is forced downwards 

 to the lower surface of the head, like that of a shark. Thus, great as the dissimilarity 

 is between the extreme forms of the snout in the species of Macncrus, there is no 

 fundamental difference in structure ; they merely represent different degrees of the same 

 line of modification. 



With regard to the scales, there is also every gradation from the small-scaled 

 Malacocephali to the large-scaled Macrunis longibarbis. In very young specimens of 

 all species the scales formed at first are always thin, without any armature, in fact 

 cycloid. Spines appear only after some time, generally in the median line of the scale, 

 singly and not in series ; scales with fully developed armature are generally not found in 

 specimens under 8 inches in length. In some species which normally possess strongly 

 spiniferous scales, individuals may occur (especially such whose skin is wanting in 

 pigment), in which the spines are much more feeble and scarcely visible. And finally, 

 there are species in which the cycloid structure of the scales remains normally persistent. 

 Thus, neither the size nor the structure of the scales can be safely used as a generic 

 character. 



The serrature or smoothness of the dorsal spine is a constant character,' the serrature 

 but rarely becoming obsolete with age, and never disappearing altogether so far as my 

 observations go. In very young specimens the barbs are comparatively further apart 

 and less numerous than in older ones, in which the barbs of more recent growth are more 

 adpressed to the body of the spine, more closely arranged, and more numerous at the 

 base than towards the top. 



The dentition is an infallible aid in the discrimination of the species, and therefore is 

 used here as the main character in their arrangement. 



1 have generically separated from the Macruri those forms which do not possess 

 the characteristic attachment of the first branchial arch to the outer wall of the gill- 

 cavity ; but with regard to the host of the other species, I do not think that they should 

 be more widely separated than in the form of subgeneric groups ; there is no doubt that 

 many more species will come to light, which may lead to a different mode of ai-ranging 

 the species. 



It is singular that, although the Macruridse are represented in the deep-sea by such a 

 great number of species and individuals, no young specimens in the Krohnius-stage'' were 

 obtained by the expedition. 



> This character has also been iisetl by Lutken in a synopsis of genera and species, Vid. Meddel. not. Form. 

 Kj()benhavn, 1872, p. 4. 



2 Gf. Emery, N. Acmd. d. Line. Rom., voL iii., 1879. 



