It is difficult to entirely free ourselves from the flattering and almost universal idea that parents are true originators or creators of copies of themselves. But the main truth, if not the whole truth, is that they are merely the transmitters of types of which they and their offspring are alike more or less similarly moulded resultants. A parent is a trustee. He transmits, not himself and his own modifications, but the stock, the type, the representative elements, of which he is a product and a custodian in one. It seems probable that he has no more definite or "particulate" influence over the reproductive elements within him than a mother over the embryo or a vessel over its cargo. Parent and offspring are like successive copies of books printed from the same "type." A battered letter in the "type" will display its effects in both earlier and later copies alike, but a purely extraneous or acquired flaw in the first copy is not necessarily repeated in subsequent copies. Unlike printer's type, however, the material source of heredity is of a fluctuating nature, consisting of competing elements derived from two parents and from innumerable ancestors. Galton compares parent and child to successive pendants on the same chain. Weismann likens them to successive offshoots thrown up by a long underground root or sucker. Such comparisons indicate the improbability of acquired modifications being transmitted to offspring. That parts are developed in offspring independently of those parts in parents is clear. Mutilated parents transmit parts which they do not possess. The offspring of young parents cannot inherit the later stages of life from parents who have not passed through them. Cases of remote reversion or atavism show that ancestral peculiarities can transmit themselves in a latent or undeveloped condition for hundreds or thousands of generations. Many obvious facts compelled Darwin to suppose that vast numbers of the reproductive gemmules in an individual are not thrown off by his own cells, but are the self-multiplying progeny of ancestral gemmules. Galton restricts the production of gemmules by the personal structure to a few exceptional cases, and would evidently like to dispense with pangenesis altogether, if he could only be sure that acquired characters are never inherited. Weismann entirely rejects pangenesis and the inheritance of acquired characters. This enables him to explain heredity by his theory of the "Continuity of the Germ-plasm."[67] Parent and offspring are alike successive products or offshoots of this persistent germ-substance, which obviously would not be correspondingly affected by modifications of parts in parents, and so would render the transmission of acquired characters impossible. INVERSE INHERITANCE.