REPORT ON THE AMPHIPODA. 553 



description of the Malacostraca, p. 447, "the head includes in all cases, behind the 

 mandibular segment on which two paragnathi form a kind of underlip, the segments of two 

 pairs of maxillae. The latter preserve more or less the character of phyllopod feet. The 

 head, therefore, consists of five segments, each with its pair of appendages, viz., two pairs 

 of antennae, one pair of mandibles, and two pairs of maxillse. It is followed by the thorax, 

 which is composed of eight segments." It may be noticed also that the eyes in some 

 Amphipoda can scarcely be called lateral, and in others are apparently altogether wanting ; 

 nor is it quite accurate to say (p. 450) that " the two eyes are always sessile, compound," 

 since in Ampelisca they are simple. 



At p. 451 the suborder Amphipoda are thus defined : — " Arthrostraca with laterally compressed 

 body, with gills on the thoracic feet and an elongated abdomen, of lohicli the three anterior 

 segments bear the swimming feet, icliile the three posterior bear jwsteriorly directed feet 

 adapted for sjmnging." 



The plates forming the brood-pouch are here called oosfegifes. " The eggs pass into tbe brood- 

 pouch and there develop. The yolk sometimes {G. locusta and other marine species) 

 . undergoes a complete segmentation. Sometimes [G. pulej;), after a superficial segmentation, 



a peripheral cell-layer is separated, which develops into a delicate blastoderm beneath the 

 egg membrane. A ventral primitive streak is then formed, and on the dorsal side, beneath 

 a differentiation which has been erroneously taken for a micropyle, a peculiar globular 

 organ makes its appearance; this is the first rudiment of the cervical gland (dorsal orijan), 

 which is confined to embryonic life. The appendages are developed from before backwards 

 on the ventrally flexed body of the embryo. The young animals usually possess at hatching 

 all their appendages and in all essential points have the structure of the adult animal, but 

 tbe number of joints of the antenna and tbe special form of the legs still present differences. 

 In the Hyperina alone the just hatched young may be without abdominal feet, and differ 

 so much in their form from the adult that they may be said to undergo a metamorphosis." 

 The following classification is made : — 



" Tribe I. — Lsemodipoda. Amphipioda with cervically iilaced anterior legs ami rudimentary 

 apodal abdomen.^' "The abdomen is small and reduced to a short protuberance destitute of 

 appendages." This statement requires modification. Caprella linearis, L., and Cyamus 

 ceti, L., are given as examples. 



" Tribe 2. — Crevettina. Amphipoda with small head, small eyes, and multiarticulate pediform 

 maxillipeds." "The coxal joints of the thoracic legs have the form of broad and large 

 epimeral plates. The abdomen has always the full number of segments. The three 

 posterior pairs of abdominal feet {uropoda) are well developed and often much elongated." 

 The epimeral plates, however, are not always large, nor are all the uropoda always well 

 developed. Three families are assigned to the Crevettina : the Corophiidte, in which "the 

 coxal joints of the legs are frequently very small " ; the Orchestiidaj, and the Gammaridse. 



" Tribe 3. — Hyperina. Amphipoda with large swollen head and large eyes, usually divided into 

 frontal and lateral eyes. Tliey have a pair of rudimentary maxillipeds functioning as 

 underlipi. 



" The antennas are sometimes short and rudimentary, sometimes of considerable size, and in the 

 male are elongated into a multiarticulate flagellum [Hypfridx). The posterior antennas may 

 in the female be reduced to the basal joint enclosing the glandular tube {Phroniyna) ; in the 

 male, on the contrary, they are folded in a zigzag, after the manner of a carpenter's rule 

 (Platyscclinse:). A paired auditory vesicle may be present above the brain (Oxycephalus, 

 Bhabdosoma)." Three families are assigned to this group, the Hyperidae, the Phronimidje 

 and PlatyscelidcTe. In the description of the family Phronimidse, the statement " Head 

 large, with projecting rostrum and large divided eye " should rather be " Head large, with 

 projecting snout or muzzle and large pair of divided eyes." 



(ZOOL. CHALL. ESP. — PART LXVII. — 1887.) XxX 70 



