204 THE CRUSTACEA 



on each of the seven free thoracic somites, the last two pairs, 

 however, being very small. If, as has been suggested above, the 

 coxal lolje of the maxillipeds is homologous with the oostegites, 

 we have the possibility of these structures being developed on all 

 the thoracic appendages. More usually, however, only the first 

 five free somites bear oostegites, and the number may be still 

 further reduced, certain genera of Arcturidae possessing only one 

 pair attached to the fourth free somite. 



Though the development of the oostegites has been traced in 

 only a few cases, it is known that important differences occur in 

 this respect. In Asellus they appear as small buds from the 

 coxopodites, increasing in size at successive moults until maturity 

 is reached. In the Oniscoidea, on the other hand, no trace of 

 oostegites is visible externally up to the moult Avhich immediately 

 precedes oviposition. Just before this moult they are developed 

 underneath the cuticle, and when this is cast off they at once 

 expand to their full size. Special structures aiding or replacing 

 the oostegites in containing the brood will be described later in 

 connection with the reproductive system, but it may be noted here 

 that in a few Isopoda (Cassidina and a few allied genera among the 

 Sphaeromidae, and some Epicaridea) the oostegites appear to be 

 entirely wanting. 



Tlie pleopods are almost always biramous, with a short proto- 

 podite in which three segments can be recognised in Bafhi/nonms, 

 and with lamellar rami generally overlapping each other with the 

 exopodite in front. One or both of the rami may be crossed by a 

 suture-line dividing it into two segments. In the simpler cases all 

 the pleopods are similar (except for the sexual modifications to be 

 described below), both rami serving as respiratory and in many 

 cases also as natatory organs. The latter function is indicated by 

 a marginal fringe of long setae and by the presence of a group of 

 coupling-hooks on the inner side of the peduncle. Pleopoda of 

 this type are found with comparatively slight modifications in the 

 Phreatoicidae, Gnathiidae, and Cymothoidae, but in some members 

 of the last two families the natatory setae, present in the young, 

 are lost in the adidt. This is the case also in the Epicaridea, where 

 the pleopoda of the adult may become much reduced or altogether 

 suppressed, or may, on the other hand, develop into arborescent 

 branchiae. In the aberrant Cymothoid Bathijnomus the endo- 

 podites bear tufts of ramified branchial filaments. Even in the 

 Cymothoidae, however, the uniformity of the pleopods is slightly 

 qualified by the fact that the endopodite of the fifth pair is always 

 devoid of marginal setae. This leads to the specialisation of 

 functions found in Sphaeromidae and Serolidae, where the anterior 

 pairs (the first three in Serolidae and many Sphaeromidae) are 

 exclusively natatory and the posterior pairs exclusively branchial. 



