But it is a colossal error to suppose that race improvement, in the strictest sense of the term, can be wrought by education.[25] The reason is simple and easily understood: Race-improvement is organic; education is extraorganic. Any change or amelioration that affects the race, the stock, the blood, must be inherited; but education is not inherited, it is not inheritable. It must be renewed generation after generation in each individual. The Sisyphus-stone of culture is rolled with infinite toil up the steep ascent by the fathers; it thunders instantly back, and must be rolled up again with equal agony and bloody sweat by their children. All must start at the same centre of ignorance, and beat out a long and arduous path to the ever-widening circumference of the farthest knowledge. The son of the learned and the son of the unlearned have equal chance side by side in the race for learning. If the children of the cultured acquire more readily than their fellows, it is not because they have inherited parental culture, but only the inherited parental capacity for culture; not because their parents knew more, but because they had more inborn power to know. Had circumstances doomed the savant to ignorance, his children would not have suffered in their ability to learn. Nay more, if devotion to intellectual pursuits has any influence at all on the native quality of offspring, as it may possibly have in extreme cases, it would seem to be more probably hurtful than helpful; for, by impairing nutrition of the germinal cells, excessive intellectual activity may induce impotence and sterility; and the fecundity of the very highly cultured seems to have suffered measurably in Europe, if not in the United States.[26] [25] To this truth, see various testimonies, pp. 149, 154, _et passim_. [26] "The tendency of human multiplication is such that the most highly cultured families tend to disappear ... Educational influences ... are superficial as compared to Hereditary causes." Franz Boas, Pro. Am. Ass. for the Adv. of Science, 1894, p. 325.