Now that we’ve stated what evolution is and is not, we need to establish the evidence for evolution.  I’ll present three different types of evidence: circumstantial evidence (evidence which fits well with the theory of evolution but does not directly prove it; i.e. evidence for which another plausible explanation may be available), direct evidence (evidence which lends directly to supporting evolution and not another theory), and experimental evidence (evidence from the lab which reproduces evolution or evolutionary adaptation).  As with the previous chapter, please feel free to chime in, whether it’s to add more evidence, to provide an alternate explanation for evidence I’ve given, to refute the evidence, or to suggest reclassification.  Additionally, if you have evidence to the contrary, let me know so I can address it in future chapters!

Please note that this evidence is not meant to be exhaustive – it’s a quick list of some of the most compelling evidence that is easily available.  If more is needed, let me know!

Circumstantial Evidence

Phylogenetics is the area of study reconstructing the evolutionary “family tree” – that is, grouping species who are most related and determining common ancestors, then finding common ancestors for those, etc., to proceed up the tree.  In the past, this was done mostly by comparing the way species looked and behaved, which could cause a lot of false starts.  For example, ants and termites appear to be very similar in behavior and appearance, but ants are actually more closely related to wasps (order Hymenoptera), while termites are related closely to cockroaches (order Blattaria).  More recently, however, philogeneticists have been able to provide more accurate and detailed reconstructions by sequencing mitochondrial genomes and ribosomal RNA sequences (the same DNA sequencing techniques used by courts to see if two people arerelated).  In essence, this is like comparing millions of slightly different copies of the same book.  The differences between copies of the books are not completely random – they follow a hierarchical structure (all copies with change A also have change B, and all copies with change C also have change B; thus we suspect change B came before either changes A or C).  These are much more useful for building the family tree.  The fact that we -can- build this family tree at all indicates that it is likely to have exist, which would imply that the species evolved through a “common descent”.  However, DNA sequencing can only be performed on extant soft tissue (DNA isn’t preserved in fossils), so we can’t conclusively fit extinct species into a family tree this way, only note where we would expect them to fit.

The universal coding of DNA – i.e. DNA across all species is written in a single “coding language” – is evident in the fact that the same strings of DNA will produce the same proteins regardless of what species it’s in: bacterium or human, snail or whale.  Completely foreign genomes can be ported into a new species (like the lab mice who glow like jellyfish).  ATP is used as the “energy currency” by all species we know about; despite the complexity of the creatures or their place on the evolutionary family tree – that aspect of metabolism seems to have evolved very early in single-cell history. Opsins (light-sensing proteins) appear to have evolved independently on separate occasions but still perform the same function.  A common source code, like common descent, is consistent with the evolutionary model, but admittedly it could also fit into some forms of a creationist model.

Atavisms are the reappearance of features lost in previous evolutionary steps (the coding gene exists but was flipped off in the species, then flipped back on by mutation in the individual).  Examples we’ve observed (courtesy of wikipedia) include hind legs on snakes and whales, hind flippers on dolphins, extra toes on horses (modern horses have one, but horse fossils going back into archaic periods have more, up to the familiar total of five), and teeth in chickens.  Examples of these are easily googleable, but let me know if you have any trouble finding them.  In humans, long canine teeth, vestigial tails, and extra nipples that follow the milk line are the most common examples.  One could conceivably dismiss these as random mutations rather than returns to a previous form, but this raises the question: why would a creator litter a species with latent genes?  At the least, it seems unlikely.

Similarly, the shapes of developing vertebrate embryos seem to be atavistic.  Though there is some initial difference in the shape of embryos, the shapes converge at a certain point, before diverging again as the species the embryos represent become more dissimilar.  This image from Mayr is a frequently-used illustration of the similarity of the shapes, but comparisons were made as long ago as the middle of the 19th century.  The similarity has led some to remark (more poetically than scientifically) that embryos seem to pass through the stages of their vertebrate family tree – fish, amphibian, reptile, then mammal.  This shared morphology indicates the likelihood of a significant shared genetic heritage.

Reuse of features is one of the stronger circumstantial evidences for evolution. Among tetrapods (four legged creatures, including mammals, reptiles, amphibians, and even some fish), the pentadactyl (five-fingered) limb construction is a common trait – even those that don’t have five fingers (like the horse and dolphins) appear to have it in their genetic and fossil background.  The common features are: a single bone in the proximal segment (upper arm), two bones in the distal segment (lower arm), followed by five series of palm bones and digits; but the features have been adapted to different uses in different creatures depending on the niche of the creature.  This is especially evident among mammals (again, examples courtesy of wikipedia –

  • In the monkey, the forelimbs are much elongated to form a grasping hand for climbing and swinging among trees.
  • In the pig, the first digit is lost, and the second and fifth digits are reduced. The remaining two digits are longer and stouter than the rest and bear a hoof for supporting the body.
  • In the horse, the forelimbs are adapted for support and running by great elongation of the third digit bearing a hoof.
  • The mole has a pair of short, spade-like forelimbs for burrowing.
  • The anteater uses its enlarged third digit for tearing down ant hills and termite nests.
  • In the whale, the forelimbs become flippers for steering and maintaining equilibrium during swimming.
  • In the bat, the forelimbs have turned into wings for flying by great elongation of four digits, while the hook-like first digit remains free for hanging from trees.

While this is consistent with a pentadactyl-limbed ancestor evolving into the creatures we know today, in a creationist model it leaves one to question why a creator would adapt structures from one creature to another, rather than building each as was best suited for their needs?

In insects (and other arthropods) you can see a similar reuse of features modified to produce a variety of shapes from the four basic mouthparts: labium, labrum, maxillae, and manidbles. See the image here for a few widely divergent examples.

The human Chromosome 2 is a remarkable piece of evidence.  Humans have 23 chromosomes; oddly enough all other hominidae (the great apes – our local branch of the family tree) have 24 chromosomes. On the face of it, this seems to undermine the idea that humans share a common ancestor with the other hominidae.  Chromosome 2 in humans is fairly long – the second longest of all 23, and there is no chromosome of corresponding length in the other hominidae.  On further examination, we see that chimpanzees have two chromosomes that we don’t (now called Chromosomes 2a and 2b), which add up to a length similar to our Chromosome 2, and the genes on their chromosomes 2a and 2b correspond to the same locations on our Chromosome 2.  Even more tellingly, though, human Chromosome 2 has some very unusual distinctions: where chromosomes typically have a single centromere near the middle and telomeres on either end, Chromosome 2 has two centromeres, and a vestigial pair of telomeres in the middle, at the fusion point!  These kind of mutation-induced oddities are to be expected in an evolutionary model, but seem an unlikely event in a creationist model.  You can read more about Chromosome 2 here: or

An endogenous retrovirus is a viral fossil within our DNA – a section of our genetic code that seems have been inserted by a virus in the distant past.  Collectively, endogenous retroviri account for as much as 8% of the human genome.  Most of this DNA is mutated junk data that cannot produce a protein, but there does seem to be involvement between some ERVs and autoimmune syndromes like multiple sclerosis.  Some ERVs also seem to be useful during pregnancy, where they appear to suppress the mother’s immune system, thereby protecting the baby from being attacked by the mother.  Some ERVs have successfully been “freed” from the human genome and reverted to retroviruses to prove that they were in fact RVs. Again, ERVs fit perfectly into an evolutionary model, but don’t seem to make sense in a creationist model – while some have use, why litter the genome with the rest, especially when they can have such negative effects? You can read more here:

Non-coding DNA, (originally called “junk” DNA) is made up of much more than just ERVs; it’s estimated to compose 98% of the human genome. In some cases, the DNA has use beside coding (telomeres and centromeres are non-coding, as are master control genes), but much of it seems to be useless – studies have shown that portions of this non-coding DNA can be removed from lab mice with no discernible affect to the creatures.  Like the ERVs (and the ERVs are a prime example of non-coding DNA), much of this code seems to be vestigial remainders of our ancestor species that produces no overwhelmingly negative effect that would cause it to be selected for editing out in mutation. Why would a creator litter our DNA with useless strands?

Direct Evidence

One of the most obvious sources of direct evidence is the fossil record, and in particular the so-called intermediate fossils (“so-called” because all fossils are intermediate to earlier and later species).  Because these deserve so much attention, I’ll save them for their own chapter.  Expect to see the fossil record for the horse, the whale, camelids, and the transition from dinosaur to bird. The key concept here is that the fossil record clearly shows the appearance and disappearance of different species, and the divergence of parent species into the modern species we see today.  There is no reasonable explanation for the fossil record in a creation theory that does not include either evolution or a deceptive creator.

Another telling source of direct evidence is the abundance of vestigial organs and vestigial features common to all complex species, but which we see most frequently in humans (because we study ourselves in greater detail than other species).  Again, these deserve special attention and will be treated in depth in the chapter dealing with Intelligent Design, but expect to see goosebumps, Jacobsen’s Organ, Darwin’s tubercle, our extra ear muscles, the plantaris muscle, our wisdom teeth, our third eyelid, the coccyx, our little toe, and our appendix show up as examples.  Vestigial organs are markers of past evolution (just as mason’s marks help us to understand how castles and cathedrals were constructed), but they also argue against an all-powerful, intelligent creator.  In some cases they would just seem to be sloppy handiwork, but in other cases (like our wisdom teeth and appendix) the vestigial organs not only have no useful modern purpose, but can cause us serious harm.

The Galapagos Finches (from which I derived my example in the first chapter), Silverswords (a Hawaiian cactus-like plant), and Drosophila (a genus of fruit fly) are all examples of adapative radiations.  Adaptive radiation is a phenomenon that occurs most frequently in the aftermath of a mass extinction or when a creature is introduced to a new environment – essentially when a species encounters unexploited ecological niches.  This is especially common in island chains (like the Galapagos and Hawaiian islands), where there are a variety of environments in close proximity; the first bird and plant species that are able to cross the ocean from their native environment adapt rapidly to fill all of the available niches. In many cases, these adaptations lead to speciation.  See Darwin’s 14 finch species (’s_finches) which are sufficiently divergent to occupy four different genera, the Hawaiian Drosphila ( at 800 species in 2 genera, and the silversword alliance ( at 50 species in 3 genera for further details of their specific details.  We can usually identify the parent species (which frequently still exists in its native location), making adaptive radiations neatly packaged synecdoches for evolution; i.e. they make good study subjects.

The Cichlid fish of Lake Victoria are a recently studied example of adaptive radiation.  The Cichlids are a family of fish that include as wide a range as tilapia, oscar, angelfish, and peacock bass across the world; iIn Lake Victoria there are at least 500 species of cichlids which, while sharing the several features common to all cichlids, span a wide variety of colors and shapes.  As in the other cases mentioned above, we’ve been able to identify the reason for the changes leading to speciation: Even though Lake Victoria is a single large lake (a -very- large lake), the different levels of clarity and light penetration in different parts of the lake favor different colors and sizes of fish. (  A separate phylogenetic study that sampled the DNA of 14 representative species identified the ancestor species as one that swam in the East African streams when  Lake Victoria was dry 12,000 years ago.

Observed adaptation and observed evolution can be as powerful a source of evidence as the fossil record.  Examples include:

The Atlantic Tomcod ( of the Hudson River has adapted a genetic tolerance for PCB, a toxin dumped in the river from 1947 through 1976 (when it was banned) which does not degrade quickly.  A minor mutation in the tomcod (a deletion of 6 pairs) that allows the fish to survive a level of the toxin that is lethal to other species already occurs in a small fraction of tomcod who swim in other waters (indicating it’s a frequent mutation), but now appears in 99% of Hudson River tomcod.  The Hudson River tomcod handle the toxin so well that there is concern over their ability to survive if the toxin is removed from the river.

Nylon-eating bacteria ( are a strain of bacteria capable of digesting byproducts of nylon production, though these chemicals are not known to have existed prior to the invention of nylon in 1935.  To quote wikipedia directly:

“In 1975 a team of Japanese scientists discovered a strain of Flavobacterium, living in ponds containing waste water from a nylon factory, that was capable of digesting certain byproducts of nylon 6 manufacture, such as the linear dimer of 6-aminohexanoate, even though those substances are not known to have existed before the invention of nylon in 1935. Further study revealed that the three enzymes the bacteria were using to digest the byproducts were significantly different from any other enzymes produced by other Flavobacterium strains (or any other bacteria for that matter), and not effective on any material other than the manmade nylon byproducts.”

This ability has been recreated in lab by exposing similar strains to the same environment; interestingly the mutations that eventually allowed the strains to adapt produced the same results, but were different than those in the original strain.

Radiotrophic fungi ( are a species of black slime mold found in the destroyed Chernobyl reactor that have adapted a method of producing chemical energy from gamma radiation.  These species do poorly in the absence of extreme radiation and this level of radiation is not found naturally, so this is suspected to be an entirely unprecedented adaptation.

The Blackcap – a European variety of warbler – appears to be on the verge of speciation.  Recently a fraction of the migratory flock (which is native to northern continental Europe) split off from the rest and developed a new habit of wintering in southern England instead of across the Alps along the Mediterranean.  These two new flocks tend not to interbreed any longer, and it is expected they’ll begin to diverge genetically in a few dozen generations.

The Hawthorn Fly seems to have recently finished speciating.  Apples are a non-native species to North America, and after their introduction in the 18th century some hawthorn flies preferred the fruit of the apple tree, while some retained a preference for hawthorn fruit.  As of now there is very little interbreeding between apple flies and hawthorn flies (4-6% incidence of hybridization), and there is significant difference in their DNA.

It’s fair wonder how to draw the line between adaptation and speciation, especially since viable hybrids are possible between recently diverged sexually-reproducing species (like the lion and tiger, or like dingos, coyotes, and wolves). Among asexually-reproducing creatures we don’t even have the metric of hybridization and have to judge purely by genetic drift.  What’s important to understand is that there -is- no hard line between adaptation and speciation.  The theory of evolution doesn’t suggest that rapid and dramatic speciation is typical, but rather that creatures who adapt to different niches will, for one of a variety of reasons, lose the ability or desire to interbreed with each other.  As breeding groups diverge, mutations that make the reproductive rounds in one group and not the other will cause genetic drift between them, and eventually hybrids will be genetically non-viable.  The species listed above are evidences at different points along that process.

Experimental Evidence

The Silver Fox experiment is a breeding program started in Russia in the 50’s, in which the fox was intentionally domesticated (bred for docility and a good temperament around humans, since this is a feature with a genetic foundation) to replicate the unintentional long-term domestication of wolves into the modern family dog.  An unexpected result of the experiment was that the domesticated foxes took on other, un-selected-for features common to dogs, like floppy ears, a curly tail, spotted markings, a reduced musk, and biannual estrus. These traits do not normally  appear appear among silver foxes. This is evidence that even when a niche selects for only a single feature, multiple distinctive features may diverge.

Sticklebacks are a popular fish among experimental biologists.  There are over forty different morphologies (shapes/colors) living in both saltwater and freshwater environments that appear to be radically different species.  They typically will not interbreed, or cannot interbreed for reasons of physical incompatibility.  However, like modern dogs with size differences, the species can be artificially hybridized and appear to be what is called a “ring species”, which (like the hawthorn fly and blackcap) is a creature on the verge – sometimes for a very long time – of speciation.  It is possible to trace the sticklebacks’ adaptive histories to a single morphology during the glacial retreat at the end of the last ice age, when they found new niches in the uncovered freshwater lakes.  One of the features of the freshwater subspecies is an increased tolerance to cold water that would kill the marine species.  By keeping samples of the marine species in successively colder freshwater, biologists were able to reproduce the tolerance of the freshwater subspecies in a marine subspecies within three generations.

Theodore Garland continues to run a long-term experiment with laboratory house mice.  To quote wikipedia directly:

“In 1993, Theodore Garland, Jr. and colleagues started a long-term experiment that involves selective breeding for high voluntary activity levels on running wheels. This experiment also continues to this day (> 50 generations). Mice from the four replicate “High Runner” lines evolved to run 3 times as many running-wheel revolutions per day as compared with the four unselected control mice groups, mainly by running faster than the control mice rather than running for more minutes/day.

The HR mice exhibit an elevated maximal aerobic capacity when tested on a motorized treadmill and a variety of other traits that appear to be adaptations that facilitate high levels of sustained endurance running (e.g., larger hearts, more symmetrical hindlimb bones). They also exhibit alterations in motivation and the reward system of the brain.”

An even more convincing experiment is Richard Lenski’s e. coli experiment, which has run past 50,000 generations ( To understand the significance, you really have to read through the link, but the gist of it is as follows: Because E. coli reproduces asexually, the only genetic changes introduced to their successive generations are through mutation. Because E. coli can be frozen indefinitely, generations at any point during the experiment can be archived to maintain a “fossil” history, which can be revived and rerun in the future.  In Lenski’s experiment there were a number of significant mutations, but the most surprising came after about 32,000 generations, when one of his colonies developed the entirely new ability to metabolize citrate in oxygen – essentially the ability to live on a new food source.  To quote wikipedia (again):

“Examination of samples of the population frozen at earlier time points led to the discovery that a citrate-using variant had evolved in the population at some point between generations 31,000 and 31,500. They used a number of genetic markers unique to this population to exclude the possibility that the citrate-using E. coli were contaminants. They also found that the ability to use citrate could spontaneously re-evolve in populations of genetically pure clones isolated from earlier time points in the population’s history. Such re-evolution of citrate utilization was never observed in clones isolated from before generation 20,000. Even in those clones that were able to re-evolve citrate utilization, the function showed a rate of occurrence on the order of once per trillion cells. The authors interpret these results as indicating that the evolution of citrate utilization in this one population depended on an earlier, perhaps non-adaptive “potentiating” mutation that had the effect of increasing the rate of mutation to citrate utilization to an accessible level (with the data they present further suggesting that citrate utilization required at least two mutations subsequent to this “potentiating” mutation).”

Unfortunately, it’s difficult to imagine a controlled experiment that would reproduce major evolutionary change in a vertebrate, like inducing a colony of reptiles to sprout feathers and become birds. Those kinds of changes require thousands, even millions of generations, which in complex creatures means thousands or millions of years (not to mention the number of creatures you’d have to maintain to provide a sample big enough for that series of mutations to have a chance of occurring and taking hold.  Fruitflies, fish, and bacteria are the evolutionary scientist’s friend, and even then the changes are subtle, like the differences between two kinds of ant, rather an ant and a wasp.  In a later chapter I’ll address this problem of experimental evidence more closely.  For now, it should suffice to say that evolution has been experimentally demonstrated in simple organisms, and it’s been observed in more complex organisms. Most of the evidence that exists, both from contemporary and historical sources, supports the evolutionary model wherever we look for it.  What does not directly support the theory of evolution does not refute it.  The same cannot be said in any meaningful way for creationism.