BEACHTUEA TEOJI TOREES STEAITS. 5 



A. Milne-Edwards in his monograph of the Cancridae {I. c. 1865) abandoned the use of 

 the character drawn from the shape of the finger-tips as a generic distinction, but he 

 retained the genus Liomera, placing it in the group of genera having the l)asal antennal 

 joint " s'unissant seulement au front par son angle interne," and separating it thus from 

 Carpilodes by a character in direct contradiction to the origmal definition of the genus. 

 Miers in 1880 {I. c.) referred our species to Carpilodes, accepting that genus in Milne- 

 Edwards's sense as having the basal joint of the antenna " produced along the exterior 

 margin of the infero-lateral frontal process so as to enter partly witliin the interior 

 orbital hiatus" (Chall. Rep., Brachyura, p. 133, 1886). Altliough thus deprived of its 

 type species, the genus Lioniem was retained by Miers, who defines it (Rep. Voy. 

 ' Alert,' Crust, p. 528) very much as Milne-Edwards liad done, and refers to it certain 

 species wliich he later included (Chall. Hep. p. 125) in the genvis Xantho. Ortmann 

 and Alcock retu.rn to Milne-Edwards's position, including this species in Liomem and 

 defining it as having the basal antennal joint not entering the orbital hiatus, and the 

 first-named author gives a figure of L. ctiictinuina to illustrate this very point. It will 

 thus be seen that of the authors who have examined this species Dana and Miers regard 

 the basal antennal joint as entering the inner orbital hiatus, while Milne-Edwards, 

 Ortmann, and Alcock state explicitly that it does not. As a matter of fact, Liomera 

 occupies in this respect an intermediate position between two extremes, which are 

 connected by a continuous series of gradations. On the one hand, we have forms where 

 the basal joint lies nearly longitudinally and meets the posterior process of the front at 

 its tip, so that the short line of junction between the two is transverse to the axis of the 

 joint ; on the other liand, we may have the basal joint lying very obliquely to the axis of 

 the body, meeting the frontal 2>i'ocess with its inner edge, so that the prolonged line of 

 junction is approximately parallel to the axis of the joint, which thus lies more or less 

 comjiletely in the hiatus between the frontal process and the low^er wall of the orbit. In 

 the present species, however, the short trapezoidal basal joint meets the frontal process 

 at its tip, but the short line of junction between the tw'o is obliquely placed w'itli 

 reference to the long axis of the joint, so that a small portion of the latter may be 

 regarded as lying in the orbital hiatus between the frontal process and the suborbital 

 ■wall. Moreover, some individual variation in this resjject is observed when a series of 

 specimens is examined, and, if we may judge by the analogous case of Actcea calciilosa 

 referred to below, it would seem that this character is liable to change with the growth 

 of the individual. 



Atergatis flouidus (L.). 



Atergatis foridus, Alcock, Jouru. Asiatic Soc. Beuji-al, Ixvii. (2) p. 98 (1898). 



Two small specimens of this widely distributed and common Indo-Pacific species 

 The carapace of the larger is 10 mm. long and IG mm. broad, the relative breadth being 

 somewhat greater than in larger specimens. 



Loc((Uty. " Murray Island, reef." 



