, BORRADAILE— ON THE PONTONIIN^ 333 



which are so widely separated, and differ even in respect of the medium through which 

 light reaches the eye. 



Genera with depressed bodies have the carapace hollowed behind the eye to give 

 room for it to be turned backwards. In Coralliocaris this depression becomes a 

 sharply-marked pit. There is here a striking repetition of the condition which in the 

 crabs may be traced through further stages till it ends in the formation of a closed orbit. 



In the antennules (Plate 52, fig. 1 c), the basal joint bears on its outer side a thin 

 flange which, generally speaking, varies in width with the body. At the base of this 

 flange, on the outer side, stands the stylocerite, a forwardly-curved, strong spine, which is 

 usually less than half the length of the joint, but sometimes, as in Harpiliopsis and 

 Coralliocaris, becomes more important. At or near its end the flange bears another 

 spine, smaller than the stylocerite. In Urocaridella and Urocaris the flange has a 

 rounded, fringed tip, projecting beyond the end of the joint, with the distal spine at 

 its side, so that the whole structure recalls the antennal scale. In PalcBmonella and 

 Periclimenes the end of the flange, slightly concave or convex, and usually fringed, 

 is nearly on a level with the end of the joint, and bears its spine at its outer angle. 

 In some species oi Periclimenes (P. frater, etc., Plate 53, fig. 6 c) there are two spines. In 

 the heavier genera there is an unfringed projection forwards which bears the spine at 

 its tip. In some species of Conchodytes this projection is blunt, the spine having dis- 

 appeared. A well-developed statocyst is always present, which is not the case in all 

 Carides. The second and third joints of the stalk vary a good deal in form. Typically 

 they are sub-cylindrical and about as long as broad. 



The inner flagellum is simple. The outer is more complex and shows variations 

 of some importance. It is always thickened at the base, and bears in the thickened 

 region a fringe of coarse, sensory hairs. At the end of this region, in all genera 

 except Typton, a cleft separates a long, slender flagellum from a short, thick one, which 

 is a continuation of the thickened region. The depth of the cleft varies. Usually it is 

 less than the length of the uncleft basal region. In most Periclimenes (Plate 52, flg. 1 c), 

 some Palcemonelloe, and the heavier genera, it is a good deal less. In some Periclimenes 

 (as P. compressus, Plate 55, fig. 18), some PalcBiuonellce (as P. hatei), and in Urocaris, 

 it is nearly or quite as long. In Urocaridella (Plate 53, fig. 2 a) and Ancyclocaris it is 

 longer than the basal region, and the antennule is then said to be " triflagellate." 

 This is the condition found in other Palsemonidse. Since the genera of Pontoniinse 

 which exhibit it are in other respects more primitive than the rest of the sub- 

 family, it would appear to be in their case an ancestral feature, though of course the 

 appearance of the third flagellum is in itself an innovation, and is one of the non- 

 primitive features of the Palaemonidse. The shortening of the cleft in higher Pon- 

 toniinse is thus a return to primitive conditions, not a retained primitive feature. It 

 is comprehensible that the lessened activity of these members of the subfamily should 

 remove the necessity for a very mobile sensory apparatus. 



The antenna (Plate 52, figs. 1 c and c') is of the normal caridean type. On the first 

 joint (coxocerite) the tubercle for the opgning of the green gland takes the form of a 

 well-developed process on the inner side. The second joint (basicerite) is divided 



