Fossil use in calibrating molecular dating

[delboy]

When DNA/RNA is used to date common ancestors I understand that fossils are needed for calibration since rates of mutation vary considerably. I know fossils can normally be dated accurately but I don't understand how they help. If a fossil is found that resembles what the common ancestor must have been like I assume it cannot be known for sure if it came before or after the split in evolutionary lines, so I don't see that that helps.

Can anyone explain?

[Arete]

A known common ancestor represents an "at least" bracket for the age of a node - as in the clade had to be in existence by at least this long ago. However, of course the origin of the clade could have occurred before the fossil, so it doesn't represent a hard lower bound for the age of the node. As a result, a fossil calibration generally allows you to put a log normal prior distribution around the node for age.