Subgenus Nannomys. Whether samples reflect only one or a complex of species is unresolved. Meester et al. (1986:283) noted that Van der Straeten "regards minutoides as a complex of different species and considers East and West African taxa different from those occurring in Southern Africa." F. Petter and Matthey (1975:3) recognized only M. minutoides, noting that among all taxa referable to that species ". . . it is still impossible to recognize those which morphologically merit specific rank." Karyotypes, however, are more revealing, and by using them F. Petter and Matthey (1975) could distinguish the typical South African minutoides (2n = 18-19), populations from all of West Africa and a part of Central Africa (2n =18-34), and populations from southern East Africa (2n = 30). The taxon grata (or gratus) is often listed as a separate species (Hatt, 1940a; Hollister, 1919; F. Petter and Matthey, 1975). Unresolved also is the geographic distributionof M. musculoides, and the nature of the biological relationship between it and M. minutoides. Samples from Angola and Zambia, for example, have not been critically studied to determine whether only one or both species are present (e.g., F. Crawford-Cabral, 1998; Petter and Matthey, 1975). The considerable chromosomal variation among samples from West Africa was documented by Jotterand (1972), Jotterand-Bellomo (1984, 1986), and Matthey (1967b) under the identification of minutoides/musculoides, a label that reflects current understanding of specific limits in this complex. Analyses of sequences from six genes indicate subgenus Nannomys, as represented by M. musculoides, is sister-group to subgenus Pyromys, which in turn phylogenetically connects with subgenus Mus (Lundrigan et al., 2002; as minutoides from Kenya).

Pertinent reviews covering taxonomy, morphology, distribution, and biology are available for populations in Senegal (Duplantier and Granjon, 1992); Gambia, Sierra Leone, and Ghana (Grubb et al., 1998); Accra Plains of S Ghana (Decher and Bahian, 1999); Nigeria (Happold, 1987); West Africa in general (Rosevear, 1969); S Sudan (Setzer, 1956); Uganda and Kenya (Delany, 1975; Hollister, 1919); Tanzania (Swynnerton and Hayman, 1951; Stanley et al., 1998; 2000, 2002); Zambia (Ansell, 1978); Malawi (Ansell and Dowsett, 1988); Dem. Rep. Congo (Misonne, 1974; specimens in AMNH); Republic of Congo (Dowsett and Granjon, 1991); Angola (Crawford-Cabral, 1998). Neither Yalden et al. (1976, 1996) nor Rupp (1980) recorded M. minutoides from Ethiopia, but we have examined many specimens from that country (in BMNH and FMNH).

The taxon kasaica (holotype from Dem. Rep. Congo, Kasaï Occidental Province, Kananga [= Luluabourg]) was originally described as a subspecies of Leggada bella (Cabrera, 1924), but subsequently treated as a species of Mus belonging to the M. sorella group (F. Petter, 1981b), an arrangement provisionally followed by Musser and Carleton (1993). We place it in the synonymy of M. minutoides after Musser’s study of the holotype graciously sent to him by Dr. Josefina Barreiro of the Museo Nacional de Sciencias Naturales, Madrid. The skin and skull exhibit all the morphological traits characteristic of M. musculoides and none of the diagnostic features peculiar to species in the M. sorella group.