SCOMBROMOHPHS. 



69 



genera, e. g. Jfolocentruni, liave never surmounted. 



Bp.ryx decadactjiJus, which attains a length of about 

 52 cm., is oni^ of the rarest of fishes, not only in Scan- 

 (lina\ia, where onh' 3 specimens have been found and 



these confined to the 



neiiilihouriK 



)od of Bergen and 



the spines of the head which generally appears with age ! stages cjf development as other, apparently youngei* 

 — the branched preorbital spines are those whicli still 

 remain longest and strongest, while the spines on the 

 nasal l)ones and on \\w forehead between the eyes al- 

 most disappear, and the spine at the lower posterioi- 

 corner of the liranehes of the lower jaw becomes in- 

 significant — we must also assign the almost complete 

 disappearance of the bony ridge on tlie opercuhim, 

 whicli in several of tiie preceding forms we have seen 

 prolonged into a spine at the hind margin. It should 

 also prove interesting to examine some young specimens 

 in order to see whether the two "sharply denticulated 

 ridges" which occur at the rounded corner of the pre- 

 operculum, do not correspond in their development to 

 some evanescent spines on the same bone. The denta- 

 tion of the scales of the body seems also to undergo 

 a completely analogous diminution as tiie iish gro\vs 

 older. Their free surface, above and below the longi- 

 tudinal groove, is fringed in youth with from 16 to 20 

 or more rows of spines, which afterwards change into 

 denticulated ridges and tinally into still smoother bars, 

 a difference which may also l)e observed to exist be- 

 tween the scales of the back and cheek and those of 

 the belh'. All these changes indicate that Beri/.r, though 

 one of the oldest generic forms and though it retains 

 the original type of the family in some respects, has 

 still i)assed throuii'h chanii'es of form which mark such 



taken at a depth of 200 fathoms, but also in the deep 

 water between the <'anar\' Islands, Madeira, the Azores 

 and Portug.il, the other locality wiiere it has been met 

 with in the Atlantic. Tiiere is a significant peculiarity, 

 howe\<'r, in connection with its geographical range, not 

 only that it thus lives in so far distant parts of the 

 Atlantic, a pi'oof of the fad that tlie deep-sea fauna is 

 uniform and made up of liard\-, ])rimitive forms, hut 

 also that it forms a connecting link between tlie fauna 

 of the Mediterranean region and the Sea of Japan. 

 DoDERLEiN and Steindachner include this species", 

 together with B. splendens among the fishes of Japan. 

 As an explanation of this find and other similar ones, 

 Gunther's* remark as to the still more remarkal)le range 

 of certain shore fishes must not be foi-gotten *": "Bold 

 as the hypothesis may appear, we can only account for 

 the singular distribution of these shore-fishes by as- 

 suminff that the Meditei'ranean and Jaijanese seas were 

 in direct and open communication with each other 

 within the period of the existence of the present Te- 

 leosteous fauna." 



SCOMBROMORPHI. 



A fusiform I/nd/j irith ii iiarmtr (loir)^ round pcdunde of flic caudal fu, a shin often, naked or only partly covered 

 with fliiu, cycloid scales^ a weak dental ecpiipment of the mouth, no denticulation of the preopercuhim in adult speci- 

 mens, u-eak spinous rays in the dorsal and anal fins and pointed {typically falciform) pectoral fins and caudal fiajis 



form the general distinctions between the series of the 

 Mackerel and Perch families. But the former, as well 

 as the latter, in contradistinction t(j the Anomalopte- 

 rous families, are stenobrachii and furnished with nu- 

 merous rays in the caudal fin (euryripidi). 



We ha\e already seen the principles on Avhich a 

 series of the Acanthopterygian Teleosteous families 

 might be grouped round the Mackerel-form as their 

 type. But there are imjiortant deviations from this 

 type within the series itself. Sometimes the body is 



naked, sometimes covered Avitii lai-ge scales; sometimes 

 it is long and low, sometimes high and compressed; 

 but these different forms are systematically connected, 

 partly l)v changes of development (gi'owtli) and partly 

 by intermediate forms even in the same family. The 

 character which may most generalh' be employed, is 

 one that in other cases whei'e the distribution is so 

 wide, is scai'cely applicable, namely, the sharp-pointed 

 form of the feeblv-sujjported fins, especially the pectoral 

 and caudal. But e\en this character must lose its 



» Denksclir. Akad. Wiss. Wien, Mat. Xalurw. (1.. Bd. XLVII, pp. ■220, iil. 



* Provided it cannot be proved tliat both these '•species" are cosmopolitan deep-sea fishes. 



' Introd. Study of FM., p. 270; JIandb. IchthyoL, p. 181. 



