The Conchoecia reticulata species-group, with descriptions of C. reticulata Muller (1906), C. caudata (1891), and two new species Celia J.Ellis Institute of Oceanographic Sciences (N.E.R.C.), Wormley, Godalming, Surrey GU8 SUB Introduction G. W. Muller (1906) described Conchoecia reticulata from Tiefsee-Expedition material. He states that the shell has a striking sculpture mainly of squares or rectangles which are organ-ised into near-horizontal rows in the dorsal half and descending rows in the ventral half, and that some animals have spines like those of C. caudata. This description seems to indicate that there were two forms of C. reticulata, a spined and an unspined form. Since this description several authors have reported C. reticulata, Granata & di Caporiacco (1949), Deevey (1968, 1980), Poulsen (1969, 1973), Angel & Fasham (1975) and Angel (1979). The material identified by Granata & di Caporiacco (1949) was undoubtedly the spined form as they used the distinctive carapace spines as one of the characters for their new genus Macroconchoecia in which they placed caudata and reticulata. Deevey (1968) did not mention the spines nor draw them, but in 1980 she confirmed that her specimens, including the specimens described in 1968, bore the distinctive C. caudata pattern of spines. Similarly Poulsen (1969) did not make clear whether the six adults caught in the Gulf of Guinea bore spines but later in his 1973 Dana Report he writes that the species is 'easy to recognise by the strong, partly spinous reticulation of the shell' thus implying that all his specimens, including those described in 1969, were spined. In a recent paper Angel (1979) was the first author since Muller to record two forms of C. reticulata. In Discovery material from 30N 23W he found a smaller deep mesopelagic form characterised by rows of short spines, a larger deeper-living form without spines, and suggested that C. reticulata was probably a mixture of two species. Further examination of Discovery material comprising 242 adult specimens led to the con-clusion that there were not two forms of C. reticulata but three. Besides the spined form and the larger unspined form noted by Angel (1979) there was a smaller unspined form. In the absence of a type specimen or syntypes, I was unable to obtain Muller's Tiefsee-Expedi-tion material, the name C. reticulata has been given to the smaller of the two unspined forms. The spined form was excluded on the grounds that it is not the species Muller figured and the terms of his description, whilst clearly including this species, implies that he regarded it as the more atypical form. The larger unspined form was also excluded as the size Muller gave for the male, 3-1 mm is outside the range of 3-4^4-0 mm for the male of this species. Furthermore the depth ranges of the adult spined form and the smaller unspined form are similar, whilst the larger unspined form lives deeper. In Discovery samples, with the excep-tion of haul 38 at station 6665 which fished a wide horizon from 1800-3600 m, the adults of the large unspined form and the spined form were not taken in the same haul. Hence it would seem more likely that Muller's unspined form would have been the small species. The large unspined form and the spined form are therefore described below with the names C. macroreticulata and C. spinireticulata respectively, re-descriptions of C. reticulata and C. caudata are also provided. Four tables giving precise station data for all material of C. reticulata, C. macroreticulata, C. spinireticulata and C. caudata identified from Discovery collections are stored in the libraries of the BM(NH) and Institute of Oceanographic Sciences. Bull. Br. Mm. nat. Hist. (Zool.) 46(4): 317-343 Issued 31 May 1984