ployed by different workers. However, where the individual samples are 

 small, the error introduced by pooling is probably no greater than the 

 error which might be imported by using such small samples to draw con- 

 clusions about the population mean. 



The limited data available suggest that siamangs have mean capacities 

 about 25 per cent greater than do gibbons. Among the gibbons, male 

 means for most species are clustered around 100 to 104 c.c, though it 

 seems that H. klossii may have a somewhat lower capacity, as Schultz 

 (1933) stressed. 



Chimpanzees. Table 5 gives available data for both species of chim- 

 panzee, Pan troglodytes being well represented, and P. paniscus poorly so. 

 Data are given for individual series and pooled totals for each sex, and 

 for combined male and female totals. The combined-sex data are in- 

 cluded purposefully, partly because several such series are available in 

 the literature, and partly because various workers had used the distribu- 

 tion characteristics of such mixed samples of extant hominoids to make 

 inferences about the distribution of capacities among fossil taxa. It was 

 therefore important to study the differences in distribution between uni- 

 sexual and bisexual samples. 



The overall male mean is about 400 c.c. in P. troglodytes and the 

 female mean close to 375 c.c. Technical and, possibly, population differ- 

 ences are suggested by the rather striking range of sample means. For 

 males, these range from as low as 381 c.c. for Schultz's (1965) large series 

 to 420 c.c. for Selenka's (1898) series. A similarly large fluctuation is evi- 

 dent among females, the sample means ranging from Schultz's low figure 

 of 350 c.c. to Selenka's high mean of 390 c.c. For both sexes, Zucker- 

 man's (1928) means are close to midway between Schultz's and Selenka's, 

 while the data of Ashton and Spence (1958) are in each sex one quarter 

 of the distance from the top of the range of means (that is, about mid- 

 way between Zuckerman's and Selenka's means). 



These regularities are reflected in a fairly constant sex ratio as de- 

 termined by the various workers. Thus, the female mean expressed as a 

 percentage of the male mean is 92.9 (Selenka), 92.7 (Ashton and Spence), 

 91.8 (Schultz), and 91.6 (Zuckerman). Only Oppenheim's male and fe- 

 male means are somewhat inconsistent, giving a sex ratio of 96.2 per 

 cent. The sex ratio of the pooled means (including Oppenheim's) is 

 93.1 per cent. 



33 U 



