Structure and growth of the baculum (os penis) in arctoid carnivores have been well described for many species. This study presents the first extensive analysis of bacular growth and variation for bears (Ursidae), based on 871 bacula of polar bears Ursus maritimus (858 of known age) that were shot in the Canadian Arctic from 1994 to 1997. Bacular length, maximal diameter and mass increased from 109 mm, 5.93 mm and 2.68 g respectively in 1-year-old bears (n = 34) to 190 (maximum 222) mm, 18.7 (24.4) mm, and 20.4 (31.9) g respectively in bears≥10 years of age (n>200). Bacular length is ∼8% of body length in adults. Bacula were moderately variable in size: coefficient of variation for length, diameter, and mass1/3 were 5.2, 10.0 and 5.4% respectively (16.3% for mass). No geographic variation in size was apparent. Bacula reached asymptotic size at 8–9 years of age. At physiological sexual maturity (∼6 years of age), bacula were at 89, 73 and 62% of asymptotic length, diameter and mass respectively (compared with body length at ∼95%). Interpretation of these findings must await behavioural and physiological information on courtship and copulation in the polar bear, and comparative data on other ursids.

A new species of horseshoe bat, Rhinolophus sakejiensis, is described from south-central Africa, near the source of the Zambezi River in north-west Zambia. A distinct combination of noseleaf, cranial and baculum characters are diagnostic of the species. It is a member of the ferrumequinum group, and its evolutionary affinities lie closest to R. clivosus Cretzschmar, 1828 and particularly the West African R. hillorum Koopman, 1989. This discovery of a new species of Rhinolophus in the clivosus complex required reappraisal of these and other related taxa. Known only from high forest in West Africa, hillorum is the closest relative of sakejiensis collected in mesic savanna in south-central Africa. Comparisons of this new species with other large Afrotropical Rhinolophus shows that hillorum is specifically distinct from clivosus, and endorses the specific status of deckeni and silvestris. The contemporary taxonomy of Afrotropical Rhinolophus was incapable of accommodating this new taxon, and the latter part of this paper argues for a more objective characterization of rhinolophids as evolutionary species.