Molecular clocks are empirical phenomena; it is hard to predict
which sequences or lineages will display clock-like variation.
The regularity of "ticks" can only be reliably determined after
the fact.
However, this is not to say that sequence comparison is invalid
as a method of deducing lineages. We do know that the time since
the last common ancestor is perhaps the best correlating factor
for predicting sequence divergence. There are also comparison
methods for developing trees that do not necessarily depend on
a clock-like rate of divergence in any particular gene. However,
dating the times of branching can get a little tricky (I suspect
that comparisons with outgroups is helpful for this work ---
John C. Avise wrote a pretty good book about molecular
phylogenies: _Molecular Markers, Natural History and Evolution_
Chapman & Hall, NY 1994. The book gives an a good summary of
the strengths and weaknesses of different methods).
Of course, the other problem with the mitochondrial eve hypothesis
that we now have evidence that part of the mitochondrial genome
may be paternally inherited (This is news for humans as a species;
in other organisms, paternal mitochondrial inheritance was previously
described). Since the original model assumed that mitochrondria were
exclusively maternal in origin, serious reworking is now required to
understand the data. I suspect that the Y-chromosome Adam models
might have related difficulties (eg. Could part of a Y-chromosome
"hitchhike" for a time on the X-chromosome? Can an XXY female
pass the Y-chromosome in exceptional circumstances?)
Regards,
Tim Ikeda
tikeda@sprintmail.hormel.com (despam address before use)