Kovács I. (szerk.): A Magyar Természettudományi Múzeum évkönyve 78. (Budapest 1986)
Taiti, S., Ferrara, F. ; Schmalfuss, H.: Chileoniscus marmoratus gen. et sp. n. from Chile (Oniscidea, Scleropactidae)
outline of the caudal part of the pleon. In some families the uropods have disappeared altogether from the dorsal position and changed to a completely ventral position, the gap between the last pleon-epimera being completely filled by the enlarged telson (Tylidae, Spelaeoniscidae, BuchnerUlo, Tendosphaera) . In some Mediterranean taxa this space is occupied by the flattened and truncated uropod-exopodite (Armadillidiidae, Atlantidium), while in all tropical conglobating families the same result has been obtained by an enlargement of the uropod-protopodite (Scleropactidae, Armadillidae , Eubelidae, Pseudarmadillidae, Irmaosidae). Phylogenetic relationships — According to SCHMALFUSS (1980) who recently discussed the phylogenetic situation of the family Scleropactidae (= Sphaeroniscidae), the following genera belong to this family: Sphaeroniscus GERSTAECKER, 1857 Scleropactes BUDDE-LUND, 1885 Spherarmadillo RICHARDSON, 1907 Circoniscus PEARSE, 1917 Neosanfilippia BRIAN, 1957 Richardsoniscus VANDEL, 1963 Amazoniscus LEMOS DE CASTRO, 1967 Colomboniscus VANDEL, 1972 Protosphaeroniscus SCHMALFUSS, 1980 while the inclusion of Sphaerobathytropa VERHOEFF, 1908 and Globannadillo RICHARDSON, 1910 into this family is considered as doubtful. SCHMALFUSS (1980) re-defines the diagnostic characters of the family and gives a cladogram of the possible phylogenetic relations within the family. In the light of the new information we propose here a slightly modified cladogram, with Chileoniscus gen. n. included (Fig. 1). The synapomorphic character of the pleopod-endopodite I with apex bent outwards, as it was assumed by SCHMALFUSS (1980), is doubtful, since there are taxa with straight tips of the endopodite not only in Scleropactes but also among the other genera. However, a sistergroup relation between Scleropactes and the rest of the family can be maintained, if the endo-antennal conglobation is considered a synapomorphic character of the latter group (in Scleropactes conglobation is exo-antennal). The specific structure of the male pereopod VII, with a lateral ridge on the merus and a corresponding groove on the ischium, is considered a synapomorphy of the Sphaeroniscusgroup containing the genera Protosphaeroniscus (see SCHMALFUSS 1980), Circoniscus (see ANDERSSON I960), Sphaeroniscus (see VANDEL 1968) and the new genus Chileoniscus (see Fig. 17 in the present paper). The genus Spherarmadillo had to be removed from this group because an examination of Spherarmadillo cavernicola MULAIK, 1960 (3 specimens, Mexico, Tamaulipas, Rancho de Cielo, cave, leg. P. Beron 31.1. 1982, Stuttgart isopod collection no. 10094) showed that these structures are not present in Spherarmadillo. If the presence of a true schism a is considered a synapomorphy of Circoniscus and Sphaeroniscus (not a very reliable character because the development of a schisma has certainly occured many times independently in conglobating isopods), the two genera can be considered as sister-groups. Another sister-group relation may exist between Neosanfilippia and Amazoniscus, taking into account the distal process on the male ischium VII. which is present in these two genera. This cladogram implies that the reduction of the 3-jointed flagellum to a 2-jointed flagellum has been achieved at least three times independently inside the family, and the evolution of a schisma has happened twice independently (in the C'irconiscus-Sphaeroniscus-gioup and in Spherarmadill o) . Other characters described in the existing literature cannot be used for the reconstruction of phylogenetic relationships since they are either plesiomorphies or autapomorphies of a single taxon.
