E. O. Powell carried out numerous experiments observing cell-generatio
n times on bacteria. His statistical methodology, though quite advance
d 40 years ago, can be enhanced by better models of dependence, by adv
anced computational methods, and by judicious use of robust methods. W
e reexamine Powell's data, focussing on his main interest in establish
ing whether there were indeed dependencies in generation times between
cells with close filial connections.