Mitochondrial DNA (mtDNA) is found outside the nucleus of the cell and is thus independent of the chromosomes. Like the y-chromosome, mtDNA is not recombined or shuffled, and it is passed more or less unchanged from mothers to their children, both males and females. Males do not pass on their mtDNA, so it can only be used to study maternal lines. There are disadvantages to mtDNA in comparison to y-chromosomes as well as some advantages.

The mutations studied in mtDNA are base-pair substitutions, where one base in the chain of DNA is replaced by another. These mutations occur much less frequently than the changes in the number of repeats in y-chromosome short-tandem repeats (STR's)--although not as infrequently as base-substitution (SNP) mutations in y-chromosomes. In general, mtDNA is useful for long-term evolutionary studies. The research publicized in the book "The Seven Daughters of Eve" used mtDNA to classify all people of European descent into seven "clans" based on long-ago matrilineal ancestors. Some people have their mtDNA tested to learn about their long-term matrilineal background (which "clan" their maternal line descends from) rather than for genealogical purposes.

Because surnames are passed patrilineally in nearly all modern societies, mtDNA is not very useful in surname studies. It can be used to try to trace descendants or suspected descendants of a known female, but this requires that the paper genealogy be complete or nearly complete first. In general the most useful genealogical purpose for mtDNA is to try to solve specific puzzles or hypotheses about unproven relationships. In my own family, two women with the last name Hamman married two Ritters (brothers of my great-great grandfather) in 1839. A reasonable hypothesis would be that these two women were sisters and the daughters of Ritter neighbor Jacob Hamman (who had four daughters or approximately the correct ages in his household in 1830). An mtDNA test of matrilineal descendants of each of the two Hamman women could substantiate or disprove the possibility that they were sisters.

Males as well as females can be tested to determine their mtDNA. This is useful if a matriline has recently come to an end or is about to end for lack of daughters For example my maternal grandmother (who had no sisters or maternal aunts who survived to adulthood) has no female great-grandchildren through females. But I and my sister's sons carry her mtDNA and could still be tested even if my mother and sister were deceased.

Unrecorded breaks in the matrilineal line of descent are less likely to occur than such breaks in a patrilineal line. Illegitimacy is rarely a problem, since there is rarely ambiguity about who was the birth mother in the same way that there may be questions about who was the biological father of a child. Adoptions of course may cause problems for both mtDNA and y-chromosome studies.

Finally, each cell contains many copies of mtDNA (usually thousands) but only one y-chromosome. DNA degrades rapidly, but the larger numbers of mtDNA make it more likely that it might be recovered in old or ancient samples. Thus mtDNA has been recovered from both Cro-Magnon and Neanderthal remains that are tens of thousands of years old, while it is currently impossible and may never be possible to recover y-chromosome DNA from such ancient samples.

Most of the major genealogical testing companies offer mtDNA tests. Rather than looking at the values of specific markers as with the y-chromosome, they sequence a portion of the mtDNA called the hypervariant region and then compare the results with a reference sequence and report the differences found.

Philip Ritter, 2005 philr@stanford.edu.