A term that has entered the popular lexicon in recent times, but which is prone to a number of mistaken ideas. So let's clarify. See mitochondrial DNA for a bit more detail, but basically the mitochondria are bits in the sex cell that are only present in the ovum, not the sperm, so you only inherit them from your mother. Mitochondria have genes of their own. So these genes represent a pure line (unsullied by sexual mixing) of descent through the maternal line. All humans have it, but males don't pass it on.

So if you analyse the DNA and track it back and work out trees of relationship you can reconstruct the female line. It tells you nothing about male ancestors. If you trace any two females back, eventually they must have a common female ancestor: the mother13 of one will be the mother15 of the other, perhaps. (More rigorous argument postponed till further down.)

By extending this argument to any group you can show that there must have been—must have been—some female who was the most recent common pure female line ancestor of everyone in that group. So, take every human being alive today. There must have been some individual female such that (i) every single one of us is her daughter's daughter's ... daughter's child; and (ii) no later female also had this property.

This woman is given the name Mitochondrial Eve.


myth: Mitochondrial Eve was "Eve"

No! She was not the first woman. She was not the first woman.

She was just one among many, nothing special about her, she had a mother and a father, she had at least two daughters and maybe some sons. (If she only had one daughter then we would call her daughter Mitochondrial Eve instead.)

myth: Mitochondrial Eve was the only woman

No! She was not the only woman. She was not the only woman.

There could have been dozens, hundreds, thousands of women alive at the same time as her, and all of whom have left descendants today.

Think of European surnames. On a desert island Smith, Jones, and Robinson have womenfolk. Generations pass. If Smith has a son, the son is called Smith. If Smith has twelve daughters, and they all marry Joneses and Robinsons, they change their name and don't pass on the Smith surname no matter how many descendants Smith has.

With European surnames they pass down the all-male line. Mitochondrial DNA passes down the all-female line. But it's the same principle. We all have plenty of ancestors from that time, but one of them happens to be the one where the pure female descent all diverges from.

myth: Mitochondrial Eve is hypothetical

No! What is speculative is when she existed. The Out of Africa theory claims that she existed quite recently, between 50 000 and 250 000 years ago. Others dispute that and say she existed way back in Homo erectus times.

It is not possible to maintain she did not exist at all. This is amenable of a probabilistic proof.

Let P be the set of all people alive today. For each person in P, take their mother. Let M1 be the set of all mothers of people alive today. Repeat. Let M2 be the set of mothers of people in M1. (Some will be in P and M1 and M2 because there are grandmothers alive today.) Keep repeating.

At every stage Mn gets smaller, because while some women can have two daughters, no woman can have two mothers. (It is possible that at some stage every woman had just one daugher, so the M's don't get smaller at that point.) Obviously the noose tightens and the set of all ancestral mothers contracts at every step. Ultimately you must get to a point where there is just one member: let's call that one Mitochondrial Eve.

Apart from this property, there is nothing particularly special about Mitrochondrial Eve. She wasn't more fertile than her fellows, she didn't have more descendants, she wasn't alone, she didn't mate with Y-Chromosome Adam... nothing. She's just a mathematical construct, who really existed.

An analogy in modern times

One more analogy: you are literate, your mother is literate, so was her mother, and so back in the female line; somewhere several centuries ago we come to the first pure-maternal ancestor of yours who could read and write. We can give her a name, say Literate Louise, even though we don't know which ancestor she was or when she lived. We can also do the same thing for your father, your father's father, and so on; and somewhere deep in your pure-paternal ancestry we have the first literate paternal ancestor: call him Literate Len. We have no idea when or where he lived either, only that there was (must have been) such a person. And of course it is highly unlikely Literate Louise and Literate Len lived anywhere near each other, in time or space.

Can male mtDNA be inherited?

An endnote with some quick thoughts on the possibility that male mtDNA gets into the zygote (see next write-up). I don't know whether this new discovery has been confirmed but for the sake of argument let's assume it turns out to be true. This would then separate our latest common female ancestor (who udneniably existed) from the common ancestor for all our mtDNA, if there was such an ancestor. But would the effect be very great?

The probability of a sperm mitochondrion getting into the zygote must be very low, otherwise it would have been observed long ago. See targeting sperm mitochondria for destruction for how it usually works. But say some proportion p survive this destruction. Then the zygote will contain all the ovum's maternal complement of mtDNA, plus a few lucky bits of male who've sneaked past. The new individual will presumably develop as a mosaic of mt genomes, with proportion p of the mitochondria in the adult being of paternal origin. Then when they make their own sex cells, p of them inherit the paternal genome. So if they're male, p of their sperm carry them, but if one of those sperm successfully reaches an ovum, its paternal mitochondria then have only a p chance of making it into the zygote.

To conclude, male mtDNA has a p2 chance of passing down into a second generation, and that's even more minuscule than the until recently undetectably small p. So I think Mitochondrial Eve is indeed the ancestor of all but negligibly much of our mtDNA.