154 DISCOVERY REPORTS 



and that of the Discovery collections from Houtjes Point, Saldanha Bay; Omer-Cooper (p. 205) 

 recorded from the Suez Canal specimens which he assigned to this species, though admitting their 

 identification to be doubtful. Except for this last record, the species has only been collected in South 



African waters. 



Genus Edotia Guerin-Meneville, 1844 



Edotia Guerin-Meneville, 1844, p. 34; Miers, 1881; Nordenstam, 1933, p. 94. 



Desmarestia Nicolet, 1849, p. 284. 



Epelys Dana, 1849, p. 426; Harger, 1880, p. 357. 



Edotea Ohlin, 1901, p. 292; Richardson, 1905, p. 394. 



Remarks. The diagnosis of this genus was given by Richardson (1905, p. 394). She stated that 

 'epimera of all the segments of the thorax are firmly and perfectly united with the segments'. 

 Richardson used the term ' epimera ' to describe the coxal plates which are developed as outgrowths 

 from the coxal joints and which may extend on to the dorsal surface and fuse with the terga of their 

 respective somites. Nordenstam (1933, p. 95), commenting on this statement, said that 'in some 

 species, however, the pereion is traversed by two lateral and parallel grooves on each side; the most 

 laterally situated of these grooves marks off the epimera. This is the case in E. bilobata and E. ocu- 

 lata.... These grooves are indistinct in E. tuberciilata, and usually they are entirely absent in 

 E. lilljeborgi, acuta, triloba, montosa, magellatiica and doello-juradoi.' Further on (p. 100), he said that 

 ' Edotia bilobata differs from the other species of the genus in having the coxal plates of the last three 

 pereion segments demarcated by very distinct suture-like grooves. These grooves or sutures are not 

 in a line with those grooves which mark off the coxal plates of the anterior segments.' I have examined 

 eight of the eleven species of this genus, four species in the Discovery collections : E. oculata Ohlin, 

 E. bilobata Nordenstam, E. oculopetiolata (Nordenstam's E. tuberculata) and E. corrugata n.sp. ; and 

 four in the British Museum collection, E. magellanica Cunningham, E. montosa (Stimpson), E. triloba 

 (Say) and E. tuberculata Guerin-Meneville; none of them conforms to Richardson's diagnostic 

 character, and in none of them do the lateral grooves mentioned by Nordenstam bear any relation to 

 the 'coxal plates of the anterior segments'. 



In all these eight species, the coxae of the limbs of the first four free thoracic segments are not 

 expanded dorsally into coxal plates. As in Synidotea, the outer margin of each coxa is ring-like and 

 (except for the first pair) is clearly marked off from the ventral surface of the pleuron by a suture, but 

 ventral coxal plates are present as extensions of the inner margins of the coxae and almost meet in 

 the middle line. In the first pair of limbs, the coxae are fused with the undersurface of the pleura of 

 the first segment. 



The coxae of the last three thoracic segments, on the contrary, are expanded dorsally into coxal 

 plates and are either fused with or separated by faint sutures from the terga of their respective seg- 

 ments ; where fusion has occurred, the position is indicated by a groove. The inner margins of these 

 coxae are also expanded into ventral plates, but these plates are fused in the middle line. 



Thus Richardson's diagnostic character is not valid for Edotia, while the demarcation of the coxal 

 plates of the last three thoracic segments, regarded by Nordenstam as characteristic only of E. bi- 

 lobata, is common to all eight species and may prove to be present in the remaining three, lilljeborgi 

 Ohlin, acuta Richardson and doello-jiiradoi Giambiagi. The similar development of coxal plates in the 

 species of Edotia and Synidotea (see p. 153) emphasizes the close relationship between these two 

 genera; this relationship is indicated by their inclusion previously in the single genus Edotia. The coxal 

 joints play an important role in the development of the marsupium or brood pouch. This is formed by 

 overlapping brood lamellae, developed as outgrowths from the coxae of some of the thoracic ap- 

 pendages. In the majority of the Valvifera the marsupium is of the normal isopod type. For example, 



