Yesterday's post discussed the method of dating by genetics. I think it is important to note that the 'molecular clock' is not foolproof, and is based on assumptions. The molecular clock was derived by estimating when two lineages split in the fossil record (via radiometric dating), and comparing the differences in genetics to come up with a rate of genetic change. Click here for Berkeley University's explanation. But this should not be looked at as precise, because it is based on the following assumptions:
Changing generation times (A mutation generally becomes fixed only from one generation to another. The shorter this timespan is, the more mutations can become fixed)
Population size (Apart from effects of small population size, genetic diversity will "bottom out" as populations become larger as the fitness advantage of any one mutation becomes smaller)
Species-specific differences (due to differing metabolism, ecology, evolutionary history,...)
Evolving functions of the encoded protein (can be ameliorated by utilizing non-coding DNA sequences or emphasizing silent mutations)
Changes in the intensity of natural selection
I thought this was worth mentioning.