Professor Karl Pearson, have desired us to accept certain conclusions
which are singularly incompatible with the opinion of their illustrious
founder, Sir Francis Galton, in favour of early marriages among those of
sound stock. By their special procedure, as rigorously critical in the
statistical treatment of _data_ as it is sweetly simple in its innocent
assumption that all _data_ are of equal value, they have proposed to
show that the elder members of a family are further removed from the
normal, average, or mean type than the younger members. This, according
to them, may sometimes work out in the production of great ability or
genius in the eldest or elder members, but oftener still shows itself in
highly undesirable characters, whether of mind or of body, the latter
often leading to premature decease. There is hence inferred a powerful
argument against the limitation of families, which means a
disproportionate increase amongst the aberrant members of the
population.

This argument really offers as good an example as can be desired of the
almost unimaginable ease with which these skilful mathematicians allow
themselves to be confused. Their inquiry has ignored the age of the
parents at marriage--or, better still, at the births of their respective
children--and has assumed that the number of the family was the
all-important point: a good example of that idolatry of number as number
which is the "freak religion" of the biometrician. Supposing that the
conclusion reached by this method be a true one--which it would need
more credulity than I possess to assert--we must conclude that, somehow,
primogeniture, as such, affects the quality of the offspring, and, on
the other hand, that to be born fifth or tenth or fifteenth involves
certain personal consequences of a special kind. Evidently we here
approach less sophisticated forms of number-worship, as that which
attached a superstitious meaning to the seventh son of a seventh son.

It seems, therefore, necessary to point out--surprising though the
necessity be--that, if the biometrical conclusion be valid, what it
demonstrates must surely be not the occult working of certain changes in
the germ-plasm, for instance, of a father, because a certain number of
his germ-cells, after separation from his body, have gone to form new
individuals (changes which would not have occurred if those germ-cells
had perished!), but rather a correlation between the _age_ of the
parents and t