122 



of the upper border of the carapace is somewhat variable and 

 this is also the case with the shape and the toothing of the 

 rostrum. The rostrum of the Scandinavian specimens (Fig. 1 — Ic) 

 usually extends just as far fonvards as the antennal scales, very 

 rarely (in 5 of the 54 specimens) it was a little shorter, reaching 

 as far as the terminal spine of the scaphocerites, in 2 niales the 

 rostrum did extend to the distal extremity of the antennular 

 peduncle and only in 3 specimens it reached 0,5 to 0,75 mm. 

 heijond the antennal scales. The proximal half of the free part 

 of the rostrum is generally slightly directed downward, the 

 distal half more or less upwards, so that the apex of the rostrum 

 in some specimens, especially in those in which the carapace is 

 much arched longitudinally, is situated below the level of the 

 upper border of the carapace, in other ones just as high as this 

 border or reaching above it. Sometimes the rostrum appears practi- 

 cally straight and in this case either slightly directed downwards, 

 like in a male long 42 mm., or slightly trending upwards, like 

 in a male long 38 mm. 



The rostral formulae of 54 specimens were the following: 



' — j — one specimen, an adult ova-bearing female. 



9 + 1 f . 8 + 2 „ 



— - — tour specimens, — - — tour specimens. 

 o o 



9 + 1 . 8 + 1 . . 



— - — one specimen, — - — eighteen specimens. 

 o o 



9 + O ^ . 8 + 1 , 



— - — two specimens, — - — three specimens. 



2 3 



9 + . 7 + 2 



— - — one specimen, — ~ — one specimen. 

 o o 



3 



9 7 + 1 



- one specimen (cT), — - — thirteen specimens. 



%■ ^ • ■ 



7 + 1 . 6 + 1 one specimen (ova-bearinff). 



— - — two specimens, — - — °^ 



