220 PROBLEMS OF RELATIVE GROWTH 



during historic times. This may be so ; but caution is indi- 

 cated. If it should prove to be the case, as is quite possible 

 on the evidence, that the various dimensions of the skull alter 

 at slightly different rates with increased absolute size, then 

 the change observed by Parsons may be merely the effect of 

 the increase of mean stature which appears undoubtedly to 

 have occurred in Europe in the past few centuries. And 

 since this is in all probability phenotypic in origin, the change 

 in skull-proportions, though of course in a sense an evolutionary 

 change, has far less significance than we should at first sight 

 be inclined to assign to it. (See Hooton, 1931.) 



Parsons himself states that the mean male stature of the 

 upper and middle classes has risen rapidly by perhaps three 

 inches to about 5 ft. 9 in., whereas the mean for the whole 

 population is only about 5 ft. 5 in., and that for the stunted 

 slum and factory populations perhaps an inch or so less. The 

 high stature of the best-nourished classes is a modern pheno- 

 menon, due to good food, exercise, etc., and was not realized 

 among the early Saxons (mean 5 ft. 6 in.). 



The change in skull-proportions to which Parsons refers is 

 one in increased relative height (taken as percentage-ratio of 

 (auricular) height in relation to total head-size — measured as 

 [length + breadth -f- height] as a standard). As a matter of 

 fact, when we make a correlation-table between relative 

 skull-height and absolute skull-size as measured by Parsons's 

 figure for [length + breadth + height], we obtain a definite 

 correlation (Table XIII). 



Naturally it would be much more satisfactory to make a 

 correlation table for all the individual measurements available, 

 but this must be a task for the physical anthropologists. 



We should also expect to find a correlation between total 

 stature and total head-size : it is noteworthy that the lowest 

 values are for poor neighbourhoods in London and for soldiers 

 in the eighteenth century ; and the educated twentieth- 

 century type, which we know to have a mean stature much 

 above that of previous centuries, has values both for absolute 

 skull-size and relative skull-height far beyond any previous 

 limit. 



As regards relative breadth (cephalic index), it seems probable 

 that different skull-types may show a change with age (size) 

 in either of two opposite directions, but here the evidence is 

 conflicting. In any case Parsons's data are most simply, 

 though not necessarily, to be interpreted as a further ex- 



