62 MR. L. A. BORRADAILE ON THE 



III. 



1. The mandible of the Palaemonidae (text-fig. 37) is deeply 

 cleft into two diverging processes, both directed obliquely towards 

 the median plane of the body. One of these — the incisor process 

 — is a thin structure, more or less ribbon-like in the Pontoniinse 

 and Desmocaris (text-figs. 32, 40), but shorter and broader in 

 Leander serratus (text-fig. 37). It trends, at its base, downwards, 

 but curves inwards and at the same time twists its outer edge 

 forwards, so that, while at its base it is nearly vertical, with its 

 width transverse to the body, at its free end it is nearly horizontal, 

 with its width longitudinal to the body. The other — the molar 

 process— is stout and subrectangular in section, and slants dorsally, 

 to end somewhat obliquely truncated on the median plane. In 

 the Pahemoninse a delicate palp, usually three-jointed, stands on 

 the anterior side of the limb, at the base of and just dorsal to the 

 incisor process, along whose outer edge it curves towards the 

 middle line of the body. The only Pontoniinse which possess a 

 mandibular palp are Urocaridella and Palwmonella . In these it 

 is two-jointed. 



The incisor process of Pontoniinaa usually ends in three teeth, 

 the midmost of which is shorter than the others, but there are 

 sometimes more. Thus in (Joralliocaris japonica there are four 

 on one mandible and five on the other, and in Conchodytes 

 tridacnce (text-fig. 40) there are on one side five, nearly equal, 

 and on the other six. In Leander serratus (text-fig. 45) there 

 are two large teeth, with on the right mandible one, and on the 

 left two, smaller intermediate teeth. In I)esm,ocaris (text-fig. 32) 

 there are four teeth on the right and five on the left mandible, 

 those at the outer ends of the row being rather larger than the 

 others. In each case the arrangement is not such that the teeth 

 of the two sides can closely interlock. The molar process of 

 most Pahemonidai (text-figs. 38-42) ends in a roughly square 

 concave surface, around which is an incomplete wall composed of 

 from four to six projecting lobes. Some of these have crescentic 

 or horseshoe-shaped rims, with their open sides towards the 

 middle of the process. Others are completely rimmed, but raised 

 more on the outer side than on the inner. In Leander serratus 

 (text-figs. 38, 39) there are four sharply distinct lobes. The lobes 

 differ a good deal in shape, and there is only a general corre- 

 spondence between those of the mandibles of the two sides. So 

 far as this correspondence goes, it is not the mirror-likeness 

 usually found in paired structures, but the two arrangements are 

 reversed, so that there is a rough sort of interlocking. In 

 L. serratus the lobes are only roughened in places. In the 

 Pontoniinse one lobe, and part of the rest of the surface, bears a 

 fur of bristles or is roughened by tubercles. This, I believe, is 

 the last remains of the clothing of bristles found on the end of 

 the molar processes of other Carides, such as the Alpheidse. In 

 the primitive Palsemonid Desmocaris (text-fig. 32) the process 



