The first two or the Lamarckian factors are the lowest in position, the most fundamental and universal, and therefore the first in the order of appearance. They precede all other factors, and were doubtless for a long time _the only ones in operation_. For, observe, all the selective factors--i. e., those of Darwin and Romanes--are conditioned on reproduction; for the changes produced by these are not in the individual during life, but in the offspring at birth. And not only so, but the operations of these factors are further conditioned on _sexual modes_ of reproduction; for all the non-sexual modes of reproduction--as, for example, by fissure and by budding--are but slight modifications of growth, and the resulting multitude of organisms may be regarded as in some sense _only an extension of the first individual_. Of course, therefore, the identical characters of the first individual are continued indefinitely, except in so far as they are modified in successive generations by the effect of the environment and by use and disuse--i. e., by the Lamarckian factors. In sexual generation, on the contrary, the characters of two diverse individuals are funded in a common offspring; and the same continuing through successive generations, it is evident that the inheritance in each individual offspring is infinitely multiple. Now, the _tendency to variation_ in offspring _is in proportion to the multiplicity of the inheritance_: for among the infinite number of slightly differing characters, as it were, offered for inheritance in each generation, some individuals will inherit more of one and some more of another character. In a word, sexual reproduction by multiple inheritance _tends to variation of offspring, and thus furnishes material for natural selection_.[16] Thus, then, I repeat, all the selective factors are absolutely dependent on sexual modes of reproduction. But there was a time when this mode of reproduction did not yet exist.[17] The sexual modes developed out of non-sexual modes. If these non-sexual preceded sexual modes of reproduction, it is evident that at first only Lamarckian factors could operate. Evolution was then carried forward wholly by changes in the individual produced by environment and by use and disuse (acquired characters), inherited and increased by integration through successive generations indefinitely. It is probable, therefore, that the _rate_ of evolution was at first comparatively slow; unless, indeed, as seems probable, the _earliest_ forms _were then_ and the _lowest_ forms _are now_ more plastic under the influence of physical conditions than are the present higher forms. Doubtless, now, in the higher animals and plants, the Darwinian factors are by far the most potent; for, among plants, where we can use these factors separately, if we wish to _make_ varieties, we propagate by seeds (sexual reproduction); but, if we wish to preserve varieties, we propagate by buds and cuttings (non-sexual reproduction).