Jean-Baptiste de Lamarck's (1830; 1914) evolutionary theory suggested a dialectic tension between the need to adapt to local circumstances and an endogenous tendency to become more advanced that was part of the alchemy of protoplasm, as it were. The endogenous predisposition drove successive generations up a ladder of ever-increasing complexity. Adaptation to local circumstances created side-branches, local deviations and evolutionary culs-de-sac. Darwin's theory eliminated the progressive tendency and explained the whole process in terms of adaptive dynamics. Darwinian organisms were agents in the weak sense that they over-produced young, which were obliged to 'struggle for existence'. There was no endogenous drive to complexify. Morphological and physiological attributes varied between organisms, and that variation, Darwin told us, was heritable.
Darwin was writing before the disciplines of genetics and ecology came into being and we must use words anachronistically to describe his work. Twentieth-century biologists used the word 'trait' to describe a heritable attribute. If we map this term onto Darwin's thesis, we can say that some types of organism (defined in terms of their traits) were better able to attract mates and more likely to thrive. The types that survived and bred were, in effect, selected by nature to transmit their traits to the rising generation. Natural selection was sufficient to explain any complexification that occurred so there was no need to invoke an endogenous tendency to become more complex. Darwin's theory of 'descent with modification under natural selection' could not work if survival were a lottery, or if variability were not heritable, or if traits were selectively neutral. His later work, as we will see, softened the last assumption, but the trait concept and natural selection remained key ideas.
Darwin needed to convince readers that the struggle for existence could never winnow populations so thoroughly that the survivors had no reserves of heritable diversity. He argued that heritable variability was omnipresent and that natural selection could multiply even tiny differences in a way that would separate the fittest from the rest. This argument placed Darwin in a bind that Stephen Jay Gould (2002) described as 'Goldilockean'. If the wellsprings of heritable diversity were to flow too strongly, then the stream of new traits could easily become a driver of evolutionary change comparable to Lamarck's progressive factor. If supplies of heritable diversity were to run dry, however, natural selection would have nothing to work with. The flow-rate had to be 'just right'.
Darwin solved his Goldilocks problem by asserting that natura non facit saltum - nature does not make jumps. Evolution was slow and continuous. There need not be much variation, certainly not enough to drive change adventitiously, because the tiniest variability between individuals could activate selective multipliers and be bulked up over countless millennia. Non facit saltum was a fudge that ensured the wellsprings of heritable diversity never ran dry or burst the banks of scientific materialism. If readers of Origin rejected the non facit saltum assumption, Darwin explained, then they must necessarily reject his theory. Thomas Henry Huxley (1863; 1864) saw non facit saltum as an unnecessary weakness and formulated an alternative salutatory model that allowed for a stick-slip dynamic.