Psychological Investigations 259

to show that there is only one mode of theocratic interference, which could upset the statistical comparison of the relative intensities of nature and nurture. He illustrates his point by supposing a caretaker tending a large number of silkworms of various breeds and Tin different ways, and that an observer watched his proceedings as well as he could, but only during the day-time and through a telescope. Now the caretaker might have a custom, of which the observer was ignorant, of feeding the silkworms in various ways during the night,"and Galton asks how this would affect the statistical conclusions. He suggests four possibilities and considers in each case that the caretaker's unobserved interference would not affect the statistical conclusions based upon classifications by nature and by nurture. But, I think, the reason of this is that Galton supposes the caretaker to pay attention in his secret proceedings only to race (i.e. nature) and to the day feedings of these races (i.e. nurture). What if he thought nothing about race or day feedings, but classified his worms by some characteristic of the individuals? Suppose he fed them differentially so as to bring all worms up to practically the same size and colour, which might be the very characteristics by which the observer had classified respectively for nurture and breed? Clearly no comparison of the effects of nurture and nature would be possible, and by less complete changes the observer might be led to very false conclusions. Further, this would be done without the caretaker knowing, as Galton supposes in his fifth alternative, that he was watched and, because he objected to being watched, devising plans to deceive the observer. There is no necessity to suppose

"the homologue would be a god with the attributes of a devil, who misled humble and earnest inquirers after truth by malicious artifice." (p. 275.)

There is in fact no need to appeal to Milton's God, who could be moved to laughter by man's quaint attempts to understand his works'.

Surely the problem is of a different kind. Either theocratic interference is perpetual and consistent, in which case it is as definite as any law of nature, and cannot be distinguished from it, and will not alter statistical results; or, it is occasional and capricious, in which case statistical samples taken under apparent sameness of physical environment will give divergent results. The general stability of statistical ratios, like the general fulfilment of prediction from so-called physical laws of nature, is the best argument against occasional and capricious theocratic interference. On pp. 277-94 Galton repeats his statistical arguments (see our pp. 115-17) against the " Objective Efficacy of Prayer." He expands to ' some extent his earlier arguments

"The cogency of all these arguments is materially increased by the recollection that many items of ancient faith have been successively abandoned by the Christian world to the domain

' Paradise Lost, Bk viii, 11. 70 et seq. :

"Or if they list to try Conjecture, he his fabric of the Heavens Hath left to their disputes, perhaps to move His laughter at their quaint opinions wide Hereafter, when they come to model Heaven, And calculate the stars."

33-2