574 MRS. E. W. SEXTON ON THE 



curious variation in the number and proportions of the joints of 

 the flagella in the same animal. This variation occurs, I believe, 

 not infrequently in the Aniphipoda,, some species of Jassa, for 

 example, having usuall}^ one joint more on one side than on the 

 other. In Grube's type specimen the primar}^ flagellum on the 

 right side has 8 joints, 7 on the left ; the right flagellum of the 

 inferior antennpe has 4, while the left has o\\\j 3 ; the accessory 

 flagella are both 2-jointed, but the one on the left antenna is 

 longer than that on the right. Grube evidently counted the 

 joints of the antennae on the left side. In two other specimens 

 examined, females of the same size, the superior antennae were the 

 same length, but in one case there were 12 joints, and in the other 

 only 9 ; the accessory flagellum in both was 3-jointed, but wdiereas 

 in the animal with 12 joints it exceeded the 2nd of the primary 

 in length (fig. 4), in the other it did not reach to the level of the 

 2nd. Figures 2 and 3 ai'e from another specimen, the right 

 accessory of which had 2 joints, the left 3. All the joints of the 

 primary, except the first 4, are provided with a sensor}^ filament 

 in addition to the setae. 



Inferior Antenna. — 4th joint of the peduncle the longest ; 

 flagellum a little longer than the 5th, subequal to the 4th in length. 

 Flagellum 3-6-jointed, the first the longest. 



First Gnathopod (fig. 5). — Hand about twice as long as broad ; 

 palm oblique, defined, as in pectinatus, by a large sensory spine 

 inset on the under siu^face ; palmar margin convex, serrate, the 

 serrations turning the corner and ending at the level of the spine, 

 it carries 5 or 6 small sharp spines submarginally on the outer 

 surface, and 8 smaller ones on the inner, all mici-oscopically 

 serrate. The finger is as long as the palm, curved, serrate, with 

 2 auxiliar}^ teeth. 



Second Gnatfiopod (fig. 6). — 2nd joint very long, equalling in 

 length joints 4-6 ta,ken together. The 4th joint, as is usual in the 

 genus, is much pi-oduced over the 5th on the inner side, and if the 

 measurements are taken along this side, the 4th, 5th, and 6th are 

 seen to be subequal to each other ; Chevreux's figiu^e (pi. v. fig. 4) 

 shows the ovxter side. The finger is about three-fifths the length 

 of the preceding joint, lightly curved, with 2 setules near the acute 

 apex. 



Perceopods 1 and 2 alike in structure, glandular, resembling the 

 tj^pe species in the shape and proportions of the joints except that 

 the 5th joint is slightly longer in proportion than in pilosus, and 

 the finger is much longer and more slender, being equal in length 

 to the posterior margin of the preceding joint. 



Hinder Perceopods very like those of the preceding species, but 

 stouter and much less elongate. Norman gives " the greatly pro- 

 duced last perseopod " as a specific chai'acter, but in most of the 

 animals examined by me the pi'oportions ai-e much as figured in 

 Grube's type specimen (fig. 1) ; in the others the 6th joint is more 

 developed. The basal joint in all is large, i-ounded oval, produced 

 downwards in a lobe, furnished anteriorly with short spines at 

 intervals and ciliated hairs, aj.id a dense cluster of the latter at the 



