A REVIEW OF THE MYSIDACEA 37 



In the female specimens the outer flagellum of the antennule is thin- 

 ner than the inner, but in the male the basal portion of this flagellum is 

 considerably thickened and is thicker than the inner flagellum. It 

 does not, however, assume the excessively stout appearance of the outer 

 flagellum in Hansenomysis and Cer atomy sis. 



The flagellum of the antenna in the female is about the same length 

 as the last joint of its peduncle and is divided into from 6 to 8 joints, 

 which are all cylindrical in shape and longer than broad. The flag- 

 ellum is, in fact, as described by Faxon in P. armiger and P. pacificus 

 and by me in P. armiger. In the male, however, it is considerably 

 longer, being more than twice as long as its peduncle, and is, moreover, 

 multiarticulate ; that is, it is composed of a large number of joints, and 

 these joints are all short joints scarcely as long as broad. There is a 

 group of very long and very delicate richly plumose setae arising from 

 the distal end of the peduncle at the base of the flagellum. 



The mouth parts of Petal ophthalmus have never been investigated. 

 I dissected these appendages from a female and figure them herewith. 

 They present no very unusual features, but are constructed on the same 

 general plan as in the Mysidacea, with minor modifications to be 

 described below. 



In the right mandible (fig. 4, a) the cutting edge consists of a 

 single chitinous ridge with one broad terminal tooth and a smaller 

 one behind it. The lacinia mobilis is present but reduced and beside 

 it is a short stout rodlike spine. The spine row is absent. The molar 

 process is well developed with three chitinous ridges separated by 

 somewhat deep grooves. In the left mandible (fig. 4, b) the cutting 

 edge is composed of two strong chitinous ridges with a narrow deep 

 groove between them. Each ridge has two bluntly pointed teeth at 

 the free end. The lacinia mobilis and rodlike spine are absent but 

 there is a single curved spine, which represents the spine row. The 

 molar process resembles that of the right side. The mandibles, there- 

 fore, differ mainly from those of typical Mysidacea in the reduction 

 or absence of the lacinia mobilis and the spine row, and, in these 

 respects, would appear to be specialized, doubtless in correlation with 

 the special development of the mandibular palp. 



The maxillule (fig. 4, c) is built on the typical mysidacean plan, 

 with lobes from the first and third joints. 



The maxilla (fig. 4, d) has the usual lobes from the second and 

 third joints but on the whole they are smaller than is usual in the 

 Mysida. The lobe from the second joint is simple and appears to 

 lack entirely the posterior setiferous expansion with its setose margin 

 such as is figured for Praunus flexuosus by Hansen {My sis flesouosa, 

 1925, pi. vi, fig. lb). The division of the lobe from the third joint 

 into two parts is barely indicated. The second joint of the endopod 



