Darwin's conception of correlation was singularly incomplete. As examples of correlation he advanced such trivial cases as the relation between albinism, deafness and blue eyes in cats, or between the tortoise-shell colour and the female sex. He used the word only in connection with what he called "correlated variation," meaning by this expression "that the whole organisation is so tied together during its growth and development, that when slight variations in any one part occur, and are accumulated through natural selection, other parts become modified" (6th ed., p. 177). He took it for granted that the "correlated variations" would be adapted to the original variation which was acted upon by natural selection, and he saw no difficulty in the gradual evolution of a complicated organ like the eye if only the steps were small enough. "It has been objected," he writes, "that in order to modify the eye and still preserve it as a perfect instrument, many changes would have to be effected simultaneously, which, it is assumed, could not be done through natural selection; but as I have attempted to show in my work on the variation of domestic animals, it is not necessary to suppose that the modifications were all simultaneous, if they were extremely slight and gradual" (6th ed., p. 226). In post-Darwinian speculation the difficulty of explaining correlated variation by natural selection alone became more acutely realised, and it was chiefly this difficulty that led Weismann to formulate his hypothesis of germinal selection as a necessary supplement to the general selection theory. The change in the conception of correlation which Darwin's influence brought about has been very clearly stated by E. von Hartmann,[359] from whom the following is taken:--"While the correlation of parts in the organism was before Darwin regarded exclusively from the standpoint of morphological systematics, Darwin tried to look at it from the standpoint of physiological and genealogical development, and in so doing he put the standpoint of morphological systematics in the shade. But the more we are now beginning to realise that systematic relationship does not necessarily imply genetic affinity the more must the correlation of parts come back into favour as a systematic principle. While Darwin only, as it were, against his will, relied on the law of correlation as a last resort when all other help failed, this law must be regarded, from the standpoint of the orderly inner determination of all organic form-change, as having the rank of the highest principle of all, a principle which rules parallel, divergent and convergent evolution" (pp. 47-8).