farther on the absolute scale above the mode than below it, a positive skew is 

 involved. ... It is probable that the slightly skewed form of curve thus deter- 

 mined is a better theoretical description of most zoological variates than is the 

 normal curve. [Simpson, Roe, Lewontin i960, pp. 145-46] 



This factor might help account for the positively skewed distribu- 

 tions in gorilla and man, but it makes all the more interesting the 

 problem of the lack of positively skewed distributions in the orangutan, 

 chimpanzee, and hylobatines. 



The consequences for Australopithecus 



The study of all available data has confirmed what the preliminary 

 investigation suggested, namely that, whatever the cause or causes, the 

 available samples of hominoid cranial capacities vary inter se in (a) sexual 

 dimorphism, (b) coefficient of variability, and (c) the presence, degree, 

 and direction of skewing of their distributions. This makes it unwar- 

 ranted, even illogical, to base a correction factor for a fossil group such 

 as Australopithecus on one or another living hominoid. In particular, it 

 is highly arbitrary to select the 2 positively skewed samples (man and 

 gorilla) to derive a correction factor that would skew the distribution 

 curve of australopithecine capacities so far to the right as to permit the 

 estimated population range to reach 680 ex. (the original estimate of the 

 endocranial capacity for the type specimen of H. habilis), as Robinson 

 has done (1966). 



It is concluded that, on grounds of variable skewness as well as 

 variable variability of available hominoid samples of capacities, it would 

 be arbitrary and of doubtful validity to select one or another skewed 

 sample of extant hominoid capacities to estimate the possible maximum 

 value in an extinct and poorly represented fossil taxon. Certainly, esti- 

 mates of maxima for Australopithecus derived in this way cannot be 

 employed validly to assist in making a decision on systematic categories. 

 For instance, it would not be valid to use information derived in this 

 manner to show that the estimated cranial capacity of H. habilis must 

 have fallen within the estimated population range of A. africanus. The 

 only formally valid estimate of the population range of a fossil taxon 

 such as A. africanus at present must be considered to be one that is 

 based upon extrapolations from the parameters of the available fossil 

 sample, rather than one that is based on inferences from the variable 



59 K 



