by the male as opposed to the female parent, let alone why offspring sometimes resembled a grandparent or some more distant relative rather than their parents. The special difficulty of accounting for the sudden emergence of monstrosities, or even less dramatically novel traits, led Darwin, in later editions of the _Origin_ as well as in _Variation,_ to become increasingly receptive to the notion that characteristics acquired by one generation might he inherited by the next. In the penultimate chapter of _Variation,_ Darwin attempted to strengthen the weak link in his chain of argument by proposing a mechanism for inheritance. He called his theory “pangenesis,” and he claimed that it explained not only ordinary inheritance—the influence of parents on their children—but also reversion, telegony, the regeneration of amputated limbs in some kinds of animals, the inheritance of acquired characteristics, and the relationship between sexual and asexual modes of reproduction and inheritance. The operation of pangenesis depended on the posited existence of unobservable units that Darwin called “gemmules,” tiny granules that were thrown off by individual cells and then circulated through the body. They had, however, an affinity for each other, which led to their aggregation in the reproductive organs or in parthenogenetic buds. They could remain latent for years, until an organism reached a certain stage of development, or for generations, until they encountered other gemmules to which they bore some special relationship. In this way a long-dormant greatgrandparental gemmule might suddenly manifest itself in a child. Since gemmules could he altered by environmental influences, they could convert acquired characteristics into the stuff of heredity. And since they were vulnerable to error, they could occasionally make mistakes, causing organs, such as limbs or tails or even heads, to develop in inappropriate numbers or in the wrong places.