I was just taking notes on this paper by Sealy and Pfeiffer (2000), and found some good quotes about body size in the Bushmen, both historically and in archaeological samples:
Historical and ethnographic sources consistently indicate that Khoisan peoples were and continue to be petite. A group of early-20th-century San studied by Dart (1937a, b) had mean statures of 155.8 cm (males) and 146.1 cm (females). Decades later, the Harvard Kalahari study found mean statures of 160.9 cm (males) and 150 cm (females). These values are comparable to the fifth centile of adult stature for contemporary North Americans (Abraham 1979). Adult weights reported for the more recent individuals are 47.9 kg (males) and 40.1 kg (females) (Truswell and Hanson 1976).
It has been claimed that environmental stressors, especially shortages of food, affected growth (Dornan 1975:80; Almeida 1965:6). The secular trend towards increasing stature among mid-20th-century Khoisan (Tobias 1978) could be seen as evidence for the influence of environmental factors.
At the same time, there is a genetic component. Low stature persists even under apparently favourable health conditions. The small body size and lean physique of living Khoisan peoples are often cited in human population biology texts as exemplary of adaptation to a hot, sometimes specifically desert, climate. Their low body-mass index is portrayed as support for Bergmann's and Allen's rules (cf. Molnar 1998, Relethford 1997). Through study of archaeologically derived materials, these hypotheses can be explored.
That's on the historic record. They examine a number of skeletons from archaeological sites and report this:
Dimensions of selected bones from the southern Cape sample are summarized in table 2. Data from one exceptionally small skeleton (UCT 345, probably a dwarf) and the three most recent skeletons with anomalous isotope values (Sealy 1997) are not included in the summary statistics for body size. The mean stature calculated from 20 male femora is 157.8 cm (s.d. = 7.9). Twenty-three female femora have a mean estimated stature of 146.9 cm (s.d. = 10.5). Greater variability among females results from some very small individuals between 4,000 and 2,000 B.P. (see fig. 4). Body size, represented by femoral head diameter to maximize sample size and divided into five sex categories, is plotted against radiocarbon date in figure 5. This figure illustrates that the smallest individuals (femora < 375 mm, therefore < 139 cm tall; femoral head diameters < 34 mm) are "probable" females, classified as female only on the basis of body size and gracility. Hence it may be inappropriate to include them in the calculation of mean female stature. Excluding the four very petite probable females, mean female stature is 149.9 cm (s.d. = 8.5). The four smallest adults appear to be of normal proportions. Only eight males and five females are sufficiently complete for living body mass to be estimated, as this requires both femoral length and bi-iliac diameter. The estimated value is 42.8 ± 6.6 kg for males, 38.3 ± 4.4 kg for females.
That's 4'10'' for females; 5'2'' for males in the archaeological sample. Bi-iliac diameter for males was 214.6 ± 16.8, for females 209.0 ± 12.3.