First, if we assume the presence of an elementary or even molecular population in a given milieu, the forms do not preexist the population, they are more like statistical results. The more a population assumes divergent forms, the more its multiplicity divides into multiplicities of different nature, the more its elements form distinct compounds or matters -- the more efficiently it distributes itself in the milieu, or divides up the milieu. [TP: 48]
[T]he phylogenesis of populations has at its disposal, in an open milieu, an entire range of relative forms to choose from, none of which is preestablished. . . . "[T]he points of reference themselves are in motion: there are only fixed points for the convenience of expression. At the level of universal evolution, it is impossible to discern that kind of reference point." [TP: 48]
Geographical areas can only harbour a sort of chaos, or, at best, extrinsic harmonies or an ecological order, temporary equilibriums between populations. [TP: 48]