Two brachyuran crabs collected at Socotra Island (Republic of Yemen) in the northwestern part of the Indian Ocean turned out to be conspecific with Sphaerocarcinus bedoti Zehntner, 1894, so far only known from a single female collected at Amboina, Indonesia, without any information on its ecology. Based on the new material and especially the availability of a male for comparison of the morphology of the copulatory appendages (gonopods), Sphaerocarcinus Zehntner, 1894, is synonymised with Caphyra Gu6rin, 1832. The extremely convex and globose carapace shape of the type specimen from Amboina turned out to be atypical and possibly caused by parasitic infection. Ecologically, C. bedoti, like most species of the genus Caphyra, lives epibiontic on soft corals (Octocorallia: Alcyonaria). During a survey on Socotra Island in the northwestern Indian Ocean in spring 1999, a pair of small crabs living on the soft coral Cladiella australis (Macfadyen, 1936) (Octocorallia: Alcyonidae) was collected by the first author (Fig. 1A). After close examination it became clear that these specimens were at least very similar to Sphaerocarcinus bedoti Zehntner, 1894, which had been described from a single female from Amboina (Moluccas, Indonesia) (Fig. lB). Subsequently no further material has ever been published, and because the original description lacks an indication of the habitat, no information on the ecology of that species is available. Zehntner (1894) established the monotypic genus Sphaerocarcinus mainly because of the very convex carapace, the truncate front, and the morphology of the external maxillipeds and female abdomen of S. bedoti, which he otherwise considered to be very close to Lissocarcinus Adams and White, 1849. Subsequently Alcock (1899) placed the genus in the portunid subfamily Caphyrinae Paulson, 1875, where he considered it to be intermediate between Lissocarcinus and Caphyra Guerin, 1832. It is, however, difficult to follow this view because he defines the genus as follows: As Lissocarcinus, but the pair of are as in Caphyra, and the carapace is very strongly convex, while on the other hand he uses the morphology of the last pair of legs (fifth pereiopods) as the main distinguishing character between Lissocarcinus and Caphyra. Obviously Alcock had not seen the type of S. bedoti and just followed Zehntner's statement regarding the similarity to Lissocarcinus without re-evaluating the morphological characters. Balss (1957) placed the genus in the subfamily Caphyrinae. Stephenson and Campbell (1960) and Stephenson (1972), in their revisionary works on the family Portunidae, did not discuss the genus in detail. Within their keys it appears to be close to Caphyra, being distinguished from the latter by the very convex carapace, the lamellate front and anterolateral border, the carpus of the fifth pereiopod being broader than the propodus, and the wrist and hand of the chelipeds being markedly carinate. Obviously the authors, however, did not examine the type of S. bedoti and did not discuss the significance of the characters they used to delimit the genus. Furthermore, the lack of a male specimen did not allow an examination of the sexual appendages (gonopods) which are known to be of high taxonomic value (Tiirkay, 1975; Guinot, 1979). Tiirkay (1971) re-examined the type specimen, providing good photographs of carapace shape, but also did not discuss the characters used for generic distinction or its systematic position. Besides being an interesting rediscovery and range extension, the specimens from Socotra, and especially the availability of