240 THE CRUSTACEA 



coxal plates of the Amphipoda must have had an independent 

 origin. The possession, by the typical Amphipoda, of a well- 

 developed inner ramus on the antennule and of a palp on the 

 maxillula are also among the characters which suggest that they 

 diverged from the common stock before the origin of the Isopoda. 

 It is quite probable, however, that their origin must be sought for 

 still further back. They do not show the late appearance of the 

 last thoracic limbs and the tendency to coalescence of the telson 

 with the last somite — characters which are common to Cumacea, 

 Tanaidacea, and Isopoda — while they possess a well -developed 

 antennal gland, which is never more than vestigial in these 

 Orders. Further, if the branchiae of the Amphipoda be, as Glaus 

 suggests, epipodites which have become shifted from the outer to 

 the inner surface of the thoracic coxopodites, it follows that the 

 Amphipoda must be supposed to have diverged, not later than the 

 origin of the Mysidacea, from the primitive caridoid stock of the 

 Peracarida. 



A regards the four sub-orders recognised in tlie classification 

 given below, it must be admitted that the differences separating the 

 Gammaridea from the CaprelliJea are not very profound. A 

 Caprellid which should unite the abdomen of Cercops with the 

 thoracic limbs of Fhtisica would be very hard to exclude from the 

 Gammaridea. The Hyperiidea hardly depart more widely from 

 the Gammaridea, but the retention of the three divisions is at 

 least convenient. The sub-order Ingolfiellidea, established by 

 Hansen for the two species of the genus Ingolfella, is not admitted 

 by Stebbing, who places the genus among the Gammaridea, but 

 there seems to be force in Hansen's contention that its inclusion in 

 that sub- order would logically require the absorption of the 

 Caprellidea also. 



The difficulties in the way of classifying the Amphipoda are of 

 a different order from those met with among the IsojDoda. We 

 have to deal with a vast diversity of forms within a comparatively 

 small range of morj^hological A^ariation, Avith the consequence that 

 genera and even families have to be established on trivial characters, 

 and their limits are often hard to define. 



Order Amphipoda, Latreille (1816). 

 Sub-Order 1. Gammaridea, Dana (1852), 



Head not coalesced with second thoracic somite ; palj? of niaxillipeds 

 with two to four segments ; all the thoracic legs jji-esent, usually with 

 well-developed coxal plates; abdominal somites generally distinct,'with 

 well-developed appendages ; eyes rarely very large. 



Family Lysianassidae. Lysianassa, H. Milne-Edwards ; Alicella, 



.1 



