348 PERCY SLADEN TRUST EXPEDITION 



/8. Rostrum downcurved, toothless, usually blunt pointed, sometimes with 

 a sharp point of abnormal fashion, usually with good ventral crest. 

 Body usually very clumsy. Third maxilliped strongly twisted. 



(1) Last two joints of third maxilliped narrow. 



Anchistus Borradaile, 1898. 



(2) Last two joints of third maxilliped broad. 



(a) No exopodites on maxillipeds. Eyes of good size. 

 Pontonides n. gen. 



(5) Exopodites on maxillipeds. Eyes small. 



(i) Dactylopodites of last three legs straight, without basal 

 prominence. 



Pontonia Latreille, 1829. 



(ii) Dactylopodites of last three legs curved, with basal prominence. 

 Conchodytes Peters, 1851. 



There can be no doubt that the most primitive genus of Pontoniinse is llrocaridella. 

 This very remarkable form has almost all the features which must have characterized 

 the earliest members of the group. The complete retention of the caridoid facies, the 

 forward direction and simple form of the legs, the gill formula only less by one small 

 pleurobranch than that of Leander or Desmocaris, the mandibular palp, the straight, 

 slender third maxilliped, the deeply-cleft outer flagellum of the antennule, all tell the 

 same tale. Yet Urocaridella can certainly not be considered ancestral to the rest of the 

 subfamily. It is specialized, probably for pelagic life, in its compression, in the peculiar 

 form of the abdomen with the long sixth segment and projecting third segment, and 

 in the long, upcurved rostrum. There can, again, be no doubt that Urocaris arose from 

 Urocaridella, through such species as Urocaris psamathe, by the loss of the mandibular 

 palp, of the gill on the second maxilliped, and of the lower teeth of the rostrum. The 

 rest of the Pontoniinse, however, must have taken independent origin from the common 

 ancestor of the subfamily, to which Urocaridella remains nearer than they, though it is 

 not transitional to them. Ancyclocaris represents such a line of independent descent. 

 By the retention, in one species, of the gill on the second maxilliped, by the narrow third 

 maxilliped, and by the deeply-cleft outer flagelhim, it is linked directly with the 

 ancestral form, for the lanceolate rostrum and some features of the abdomen make 

 a descent through Urocaridella impossible, but it has lost the mandibular palp, and its 

 stout body and legs, with the curved dactylopodites, show an interesting parallel to 

 Harpiliopsis and allied genera. Probably, however, this is due only to a somewhat 

 similar mode of life. The hump on the back of the female is a peculiarity which does 

 not recur in the subfamily. 



The remaining genera probably represent a third line of descent, through a form 

 resembling Palcemonella and Periclimenes, from which Periclimenes has departed in 

 the loss of the mandibular palp and Palcemonella in the loss of the vestige of the gill 



