50 MR. L. A. BORRADAILE ON THE 



tliis is succeeded by a sixth and a seventh, while in Cerataspis 

 (text-fig. 15) the endopodite bears four endites, making eight 

 altogether, besides the apical lobe in the position of the dactylo- 

 podite. In the last two cases it is clear that the endopodite 

 represents the whole distal part of the limb, and not, as has been 

 supposed, the fifth or sixth endite only. 



For a detailed comparison of the jointed stem formed by the 

 protopodite and endopodite of the biramous limb with the endite- 

 bearing axis of the phyllopod appendage, it is necessary to find 

 some fixed point of correspondence between members of the two 

 series. It would be natural to look for this at the base of the limb, 

 bnt the development of Penceus and other Decapoda,, during which 

 the basal joint of the appendages of the thorax recedes into the 

 side of the body, is a warning against the use of such a method. 

 On the other hand, the differences in the total number of endites 

 in different limbs makes it useless to stiii-t a comparison from the 

 distal end. The epipodites vary in numl)er, and their homologies 

 are doubtful, so that their points of attachment afford no reliable 

 clue. There remains the flabellum-exopodite. It is unfortunate 

 that the uncertainty which exists in regard to tlie homologies of 

 the endites of Branchiopoda makes even the seat of the flabellum 

 a doubtful point in that group, but, as will be seen, what can 

 be stated on this point is not without its uses in confirming the 

 evidence of other classes. In the Anostraca (text-fig. 4) the 

 flabellum is probably absent. In Conchostraca (text-fig. 5) it 

 stands opposite the junction of the second and third endites of a 

 series of six, but, quite possibly, these two members represent the 

 third and fourth of the full series, the second, which is present 

 in the Anostraca, being in Notostraca vestigial and in Concho- 

 straca either lost or fused with the third to form a single 

 elongate lobe. In Notostraca (text-fig. 3), where the whole axis 

 of the limb is more or less strongly segmented, the flabellum 

 stands on the segment of what may be the third or the fourth 

 endite. In the maxilla of the Malacostraca (text-figs. 14-18, and 

 47) the situation is clear. Here the exopodite arises opposite the 

 third and fourth endites, upon a. region of the limb which may be 

 undelimited, as in most Decapoda, but which is often marked oft' 

 more or less completely to form a segment, as in Leptosti-aca 

 (text-fig. 14), Mysidacea, Cumacea, Euphausiacea, and the decapod 

 larva Anisocaris (text-fig. 17). Where there is no exopodite, 

 as in Isopoda, Amphipoda, and Stomatopoda * (text-fig. 19), a 

 segment with two endites is found in the same position on the 

 maxilla as on that of the orders in which the limb bears an 

 exopodite. The maxilla of Ccdanus (text-fig. 13) agrees remark- 

 ably in structure with that of Malacostraca. Though it has no 



* III the maxilla of tlie Stomatopoda, tlie basipodite is represented upon the 

 outer side of the limb by a single sclerite, but the segment of the third endite is 

 diffei-entifited trom that of the fourth by the possession of a sclerite of its own upon 

 its hinder face. Tiie structure of this rather puzzling limb is elucidated in text- 

 fig. 19. Jt is interesting to note that the maxillary gland opens ou the second 

 joint, like the auteunary gland of Mysidacea and Amphipoda. 



