The Constant That Minimises Squared Error

You must summarise a dataset with a single constant $c$, used as the prediction for every observation. The dataset is

$[\, 2, \; 4, \; 6, \; 8, \; 30 \,]$

You are scored by the sum of squared errors

$L(c) = \sum_{i=1}^{5} (x_i - c)^2$

and $c$ may be any real number.

Which constant minimises $L(c)$, and what is the minimum value of $L$ itself?