BIHANG TILL K. SV. VET.-AKAD. HANDL. BAND 26. AFD. IV. N:0 S. 29 



of three times as large specimens of the other species. This 

 shortness of the head is mostly due to the short postorbital 

 region, which is less than a fourth of the distance from snout 

 to dorsal fin and consequently proves that the fish belongs 

 to the subgenus Clupea and thus offers the most valuable 

 character for distinguishing this species. The ventral fins 

 are situated more posteriorly than in the other Caspian spe- 

 cies, which on the other hand is a juvenile character. The 

 maxillary is shorter than in other species. If the relative 

 measureraents of this fish are compared with those of our 

 two species of true Clupea it is evident that it resembles 

 G. sprattus in one of the most important characters. The 

 base of the anal fin is, namely, considerably longer than ^/4 

 of the distance between snout and ventral fins. But, on the 

 other hand, the distance of the ventral fins from the snout 

 is rather large as already has been said and always con- 

 spicuously larger than the distance of the dorsal fin from 

 the snout. This easily separates C. delicatula from C. sp)rat- 

 tu.f. Smttt points out that C. sprattus differs from the Medi- 

 terranean C. phalerica in having a smaller head and a less 

 high caudal peduncle. In the former of these characters C. 

 ddicatula resembles pJuilerica, but with regard to the latter 

 C. sprattus. 



Ripe specimens of C. delicatula have been observed by 

 BoRODiNE in the Volga delta, but I should think that this 

 form is the most marine of the Caspian herrings and may 

 also propagate in the sea. 



The three remaining species which grow to a much larger 

 size all belong to the subgenus Älosa, which can be seen from 

 the anterior position of the dorsal fin the distance of which 

 is, from the snout, less than four times the postorbital length. 



They represent in their development three different bio- 

 logical types. Clupea caspia with about 120 gillrakers is a 

 plankton feeder which, as a rule, has no mandibular teeth. 

 Clupea Kessleri with about 60 — 80 gillrakers seems to feed 

 mostly on a little larger erustacea (mysidacea, cumacea and 

 amphipoda) and has teeth, although not so strong as those 

 of the follovving. Clup)ea SaposlmikoiH is a rapacious form 

 with strong teeth but only 24 — 38 gillrakers. It preys chiefly 

 upon small fish Ätherina, young herrings, gobiids (among 

 others Bcntliopjliilus) etc. but sometimes are also cumacea and 



