342 PERCY SLADEN TRUST EXPEDITION 



the minute organisms which must form the food of some species, and it is hard to imagine 

 that a Conchodytes, for instance, can have many enemies against which such weapons 

 would be of any avail. They may have sexual uses, but rarely differ in the sexes. 

 The peculiar features which they present in the several species are quite unexplained, 

 and the study of this limb in relation to the habits of the species presents an interesting 

 field of observation. 



The legs of the last three pairs are similar in all but minor details. In Urocaridella 

 and some species of Urocaris they are directed forwards as in Palcemon, but in Urocaris 

 psamathe there appears a change in their position which increases as the build of the 

 body becomes heavier, and is probably to be regarded as an adaptation to walking. 

 The basipodite is here curved so that, whereas the coxo-basipodite joint is transverse 

 to the main axis of the body, the basi-ischiopodite joint tends to become longitudinal, 

 and the limb is thrown outwards, and thus the body has a wider area of support. A row 

 of movable spines is often found on the lower side of the propodite (Plate 52, fig. 1 m), 

 and sometimes there are spines on other joints. The dactylopodites vary greatly. 

 In the more lightly built species they are generally slender, nearly straight, and simple 

 (Plate 52, fig. 1 m). In Harpilius, Harpiliopsis, Coralliocaris, and Conchodytes, they are 

 strongly curved. Coralliocaris (Plate 56, fig. 23 m) has at the base of the joint on the 

 under side a curious prominence which is usually more massive than the terminal claw. In 

 Stegopontonia this prominence is paired. Conchodytes (Plate 57, fig. 26 m) has a swelling in 

 the same position. In members of various genera the end-claw may be duplicated by a 

 spine underneath it, so that the limb becomes " biunguiculate " (Plate 55, fig. 19 and 

 Plate 57, fig. 26 m). In some cases the end-claw is exceedingly sharp, in others it is stouter 

 and blunter. No doubt each of these peculiarities is an adaptation to the substratum 

 upon which the species possessing it moves. Most of them are not as yet susceptible of 

 explanation, but when the claws are both sharp and curved they are often dug into the 

 soft tissues of the host. 



The abdominal limbs (Plate 52, fig. 1 and Plate 53, figs. 2 o — o") are of the ordinary 

 caridean type and show few remarkable features. The appendix interna is present on the 

 second to fifth pairs in both sexes, and on the second pair of the male there is an additional 

 process beside the appendix. The endopodite of the first limb is always smaller than 

 that of the others, and its inner edge is often crumpled over. Urocaridella is remarkable 

 for having an appendix interna on this limb. This feature occurs also in Am^phipalcBmon 

 and is one of many indications of a relationship between the Anchistioididse * and the 

 Palsemonidse. 



The fullest gill formula found among the Pontoniinse is that of Urocaridella and 

 Ancyclocaris aberrans. Here there is a pleurobranch for each leg, an arthrobranch for 

 the third maxilliped, and a podobranch on the second maxilliped (Plate 53, fig. 2 h), with 

 a mastigobranch on each of the maxillipeds. A tiny lobe above the base of the mastigo- 

 branch of the second maxilliped is perhaps the last vestige of an arthrobi-anch. A second 

 group of genera, comprising Urocaris, Periclim^enes, Harpiliopsis, Coralliocaris, Palce- 

 monella, Pontonides, PericlimencBus and Anchistus, have lost the podobranch of the 



* See p. 405 in this volume. 



