Small dependence of the gradient on the rate of morphogen production has been used as a criterion of biological relevance (Eldar et al., 2003). There are however documented cases of morphogen readouts not being buffered against changes in morphogen production rates (Grimm and Pflugfelder, 1996, Zecca et al., 1996, Staehling-Hampton et al., 1994). In this study, gradients were evaluated with the “$\eta$” criterion proposed by Lander et al. (2002), which provides a simple measure of how close to linearity they are (as it would be difficult for rapidly-decaying gradients to provide a useful range of concentrations over a high number of cells).