SIIKA TFISII. 



()97 



lateral etimioid lujiies. Tlie tongue is wanting. The 

 unilniiiil lione is of a singular t'orni. Its anterior, siior- 

 ter jiart (curijus) resembles a parallelo|)i[)(!(l. Its posterior 

 part is e\pan<icil into a rli<>inl>ic, iiorizontaj disk, willi 

 the two ;uiUTi<ir sides curNed upwards and the two 

 postei'ior sliar|i, hut deeply eoiica\e. Al)o\e this disk, 

 and at right angles to it, rises a strong, ele\"ated ridge, 

 from tiie extreme beginning of the bone. 



Tile form of the dorsal fin is especially characteris- 

 tic of the Sheatfish and reminds us sti-ongly of the 

 adipose fin in our Salmonoids, resembling an obliquely 

 linguit'iirm flap on a faii'l\- narrow stalk, snapped, as 

 it wcu'e, and iient backwards at the middle. The tin 

 lies at a distance from the tip of the snout that mea- 

 sures in \ oung specimens about 27 %, in old about 23 

 % of the length of the body", or in tlie former about 

 69 %, in the latter about 63 ?» of the distance l)etween 

 the anal tin and tlie ti[i of the snout. Its four inter- 

 spinal bones coalesce into one supporting Ikmk', the lower 

 extremity of which is attached between the tops of the 

 fifth and sixth neural spines (belonging to the sixth 

 and seventh vertebra-). Tbe first ray is simple but 

 distinctly articulated, the posterior two'' or three rays 

 are branched and articulated. 



The anal fin, the length (base) of which usually 

 measures more than half (about .53 or .'J4 "»'') of tiie 

 length of the body, is of almost uniform height through- 

 out its length, though its height increases behind in the 

 same proportion as the depth of the body decreases. 

 Its greatest height (longest ray) is as a ride about '/le 

 — ■' ,3 of the length of the body, but may sometimes 

 rise to about ' ^ thereof. The fin begins just behind 

 the vent with its fairly large genital papilla, at a di- 

 stance from the ti]i of the snout that measures in j'oung 

 specimens about 40 — 43 %, in old about 36 % of the 

 length of the body. Its last ray is united by a fin- 

 meml)rane to the lower margin of the caudal fin, as 

 tar as tlie black colour extends over its base. 



The caudal fin, which during life seems always to 

 be directed downwards, sometimes almost at a right 

 angle to the longitudinal axis of the bodv', mav, how- 



ever, be raisi'd and brought without \iolence into the 

 position which it occupies in our figui-c. In this posi- 

 tion, however, tiie skin forms a protuberance at the 

 dorsal margin just in front of the fin. The length of 

 this fin generally measures in j^oung specimens about 

 11 %, in old about 9 — 8'/o % of that of the body; but 

 the variations in this respect are quite as great as in 

 the case of the height of the anal fin, and in a speci- 

 men 1 metre long we have found the length of the 

 longest caudal rays to be 14 '/^ % of that of the body. 



The pectoral and ventral fins are the only ones in 

 whose various relations we apparenth- find jierceptible 

 external differences between tlie sexes. These fins are 

 relatively larger in the males than in the females, to 

 judge by our measurements of two males respective!}' 

 about 1 and 2 metres long and two females respectively 

 57 Vo <iiid 73'/o cm. long, our only specimens in which 

 the viscera were sufficiently well preserved to enable 

 us to decide the sex with certainty. 



The pectoral fins are broad, but oblicjuely oval. 

 Their insertion is partly covered at the upper angle by 

 the broad branchiostegal membrane. Their first (upper- 

 most) raj- is strong and spinous, though articulated at 

 the tip, and forms a weapon which in old specimens 

 develops a number of spines on the inside of the outer 

 (distal) part. When this weapon is employed, the spi- 

 nous rav is erected in an outward direction, a creaking 

 sound is heard, a sound wdiich the fish probably uses 

 to inspire terror, and the articulation is locked, so that 

 the ray cannot lie forced back. This is a faculty espe- 

 cially common among the fishes of this family, and the 

 manner in which this result is attained is as follows. 

 The base of the spinous ray is not only furnished as 

 usual with two articular knobs — one for each of the two 

 articular surfaces which lie on the hind margin of the 

 firm osseous connexion between the scapular and cora- 

 coid parts of the shoulder-girdle, and of which the outer 

 (the one nearer to the clavicle) is an obliquely-set and cir- 

 cularly hollowed groove — it is also furnished on the out- 

 side with a tumid, pulley-shaped, articular knob, which 

 fits into a corres])onding ear-shaped groove on the in- 



" To judge by these clianges of growth, as well as h\ the vomerine cards of tceUi, which in Ihu Indian Silurus ivynaadensis (Day, 

 Fish, fiiil., p. 480, pi. CXI, fig. 6) are separated from each other, this Indian form seems to stand nearer the original type common to both 

 species, though we must not forget that the Indian species is known only in speciinens at most 12 in. {30,5 mm.) long. Besides, external 

 differences of se.x are unknown in these species. 



' In old specimens we have only found two branched rays in the dorsal fin. 



' According to Kkoyeu, however, sometimes (in young specimens) 49 or 50 %. 



** This downward curvature of the caudal fin and its union to the anal fin are given by Sac.evkhi, (Morphol. Jahrh., X 1884), p. 5) 

 as one of the proofs of (he close relationship between llie Gyiunotoids (the so-called Electric Eels) and the Sihiroids. 



