In the 1970s, Cal Tech geneticist Edward B. Lewis discovered that by carefully breeding three mutant strains he was able to produce a fruit fly in which the balancers were transformed into a second pair of normal-looking wings.
At first glance, this might seem to provide evidence for Carroll's claim that small developmental changes in regulatory DNA can produce large evolutionary changes in form. But the fruit fly is still a fruit fly. Furthermore, although the second pair of wings looks normal, it has no flight muscles. A four-winged fruit fly is like an airplane with a second pair of wings dangling uselessly from its tail. It has great difficulty flying or mating, so it can survive only in the laboratory. As evidence for evolution, a four-winged fruit fly is no better than a two-headed calf in a circus sideshow.2
Disabled fruit flies with extra wings or missing legs have taught us something about developmental genetics, but nothing about evolution. All of the evidence points to one conclusion: no matter what we do to a fruit fly embryo, there are only three possible outcomes-a normal fruit fly, a defective fruit fly, or a dead fruit fly. Not even a horsefly, much less a horse.3
According to Peter Raven and George Johnson’s 1999 textbook, Biology, “all evolution begins with alterations in the genetic message… Genetic change through mutation and recombination [the re-arrangement of existing genes] provides the raw materials for evolution.” The same page features a photo of a four-winged fruit fly, which is described as “a mutant because of changes in Ultrabithorax, a gene regulating a critical stage of development; it possesses two thoracic segments and thus two sets of wings.”
…
Adding to the confusion, textbook accounts typically leave the reader with the impression that the extra wings represent a gain of structures. But four-winged fruit flies have actually lost structures which they need for flying. Their balancers are gone, and instead of being replaced with something new have been replaced with copies of structures already present in another segment. Although pictures of four-winged fruit flies give the impression that mutations have added something new, the exact opposite is closer to the truth.4
The probability of dust carried by the wind reproducing Dürer’s “Melancholia” is less infinitesimal than the probability of copy errors in the DNA molecules leading to the formation of the eye. 9