tion estimate of the upper and lower limits of cranial capacity in A. afri- 

 canus. 



For the robust and hyper-robust australopithecines, the data are too 

 scanty to permit a similar computation to be made. All that can be said at 

 the moment is that the 2 available estimates, 500 ex. and 530 ex., fall 

 within the A. africanus sample range and, a fortiori, within the estimated 

 population range for A. africanus. Thus, as yet there is no acceptable evi- 

 dence to support the idea that the robust australopithecines were signifi- 

 cantly larger-brained than the gracile ones (Figure 13).* 



Are we justified in using 3 s.d.s above the sample mean as the up- 

 per limit? For a normally distributed variate, it would seem to be perfectly 

 reasonable to employ 3 standard deviations above and below the sample 

 mean for estimating the limits of the population range. In fact, for many 

 variates, palaeontologists and mammalogists are content with 2.5 S.D.s de- 

 parture from the mean. Hence, in employing 3 S.D.s one might be thought 

 to be allowing an appreciable safety margin. However, it has been ques- 

 tioned whether, in the instance of cranial capacity, these limits do provide 

 a realistic picture of the population distribution. Robinson, in particular, 

 has questioned this assumption. He states: 



The mean (cranial capacity for 6 specimens of A. africanus) was 430 cm. 3 and the 

 range, using limits of 2.5 times the standard deviation on either side of the mean, 

 was 300-550 cm. 3 On comparing these figures with similar ones calculated for 

 pongids and modern man from data, based on substantial samples, provided by 

 Ashton and Spence (1958), it became obvious to me that even limits of three times 

 the standard deviation seriously underestimate the upper limit of the observed 

 range for both the gorilla and man. This being the case, it is possible that the 

 Australopithecus range could similarly be underestimated. Using the percentage 

 for man as a basis for adjustment, a corrected upper value of 680 cm. 3 was ob- 

 tained and one of just more than 600 cm. 3 on the basis of the gorilla discrepancy. 

 [Robinson 1966, p. 958] 



Unfortunately, once again Robinson has not given the essential data 

 on which he has based these computations. One would wish to know what 

 value he employed for the S.D. of man and gorilla; what upper limit he 

 employed for the ranges for man and gorilla in terms of number of cubic 



•Based on Holloway's (1970b) estimates, the robust australopithecines with a mean of 530 c.c. 

 (n = 2) are significantly greater in capacity than the gracile australopithecines with a mean of 

 442 c.c. (n = 6). 



29 K 



