Figure 4. Test of rescaling hypothesis.

Shown are the two fitted log-normal curves for flagellates and ciliates, each multiplied by the distribution median $M$, plotted versus speed normalized by $M$. The distributions for show remarkable similarity and uncertainty of estimation.

Figure 4—figure supplement 1. Data collapse as in the main figure, but using the mean speeds $U^{*}$ instead of the median $M$.

A similar quality of data collapse is seen.