Darwin's Precursors and Influences

7. Heredity

by John Wilkins

Copyright © 1996-2003

[Last Update: 21 February 2003]

Previous

Contents

Next

From the beginning, Darwin was concerned about the source of variation in a species on which selection could act, and in his notebooks sometimes played with a view of heredity not unlike modern population genetics¹. However, he had a serious problem in the way in which he eventually developed his view. This is called the problem of blending inheritance.

It was then widely accepted that use of a trait would strengthen its inheritance, while disuse would make it less strong in descendents. Sometimes Lamarck is credited (or blamed) for this opinion, but versions of it were common throughout the seventeenth and eighteenth centuries and possibly back as far as classical times and the Bible, up until the rediscovery of Mendelian genetics in 1900. Darwin accepted that use and disuse would affect heredity, and thus provide the source of variation for selection to act upon.

Darwin thought that particles called "gemmules" would move from the extremities where an organ was, back to the sex cells and, in modern terms, "reprogram" the sex cells with the new information. Had it been true (there is no logical problem with this theory, merely empirical ones) it would indeed have provided significant evolutionarily variation. Darwin's theory is called pangenesis and the factors that passed the hereditable information on he called pangenes, which was later shortened to "genes" after Mendelian genetics were developed.

However, Darwin had a more fundamental problem. He thought that on sexual recombination of pangenes, a pangene for, say, great height and a pangene for shortness would blend together to create a pangene for intermediate height. It was pointed out by a critic Fleeming Jenkin (1867), and reiterated by the anti-selectionist Mivart (1871), who was still a transmutationist, that this meant that variation would be "swamped". Fisher in 1930 calculated that with blending inheritance roughly half of all variation would disappear each generation, while no more than 1/1000 of variation now in existence in any species could be more than 10 generations old (or 20 if you take sexual recombination into account). Consequently, the rate of inherited novelty (what we now call mutation) had to be extremely high. To accommodate this, Darwin had to rely on the use-and-disuse theory. When Mendelian genetics flowered at the turn of the century, this problem led a number of the new geneticists to argue that Darwinian evolution (i.e., natural selection) was dead. Fisher and Sewall Wright in the 1930s showed that it wasn't, and that indeed, Darwinian selection flowed as a logical consequence of the results of the new genetics. On a Mendelian picture, genetic variation can persist in a population for a very long time, which increases the likelihood that a new mutation will meet copies of itself and thus permit new traits to become fixed.

There is little point looking for Darwin's precursors to this: it was commonly held that heredity worked in this way². Sometimes, this view of heredity is called Lamarckism, because in the late 19th century a group of non-Darwinian evolutionists took to calling their own views Neo-Lamarckism, meaning that genetic novelty was not random. Echoes of this view persisted until their final defeat in the 1950s, and remain defeated despite such phenomena as meiotic drive and extra-chromosomal mechanisms of heredity. However, it was not original with Lamarck, although he made his own contributions to it.

Note, though, that Darwin was very cautious about his statements on the import of the mechanism he proposed for the theories of evolution, often referring to the abysmal ignorance of the science of the day³. What Darwin needed for selection to work was persistant variation, and this was not a deduction on his part but observation in the wild, to which he devoted two chapters (chapters I and II) in the Origin.

Previous

Contents

Next