+
+Curvature is a measure of the intrinsic propensity of specific lipids
+to form micelles (positive curvature), inverted micelles (negative
+curvature), or planar sheets (zero curvature). In this formalism,
+curvature is measured as the ratio of the size of the head to that of
+the base, so negative curvature is bounded by $(0,1)$, zero curvature
+is 1, and positive curvature is bounded by $(1,\infty)$. The curvature
+can be transformed into the typical postive/negative mapping using
+$\log$, which has the additional property of making the range of
+positive and negative curvature equal, and distributed about 0.
+
+As in the case of unsaturation, void formation is increased by the
+presence of lipids with mismatched curvature. Thus, a larger
+distribution of curvature in the vesicle increases the rate of lipid
+insertion into the vesicle. However, a species with curvature $e^-1$
+will cancel out a species with curvature $e$, so we have to log
+transform (turning these into -1 and 1), then take the absolute value
+(1 and 1), and finally measure the width of the distribution. Thus, by
+using the log transform to make the range of the lipid curvature equal
+between positive and negative, taking the absolute value to cancel out
+exactly mismatched curvatures, we come to an equation with the shape
+$x^{\mathrm{stdev}\left|\log cu_\mathrm{vesicle}\right|}$, and a
+convenient base for $x$ is 10, yielding:
+
+{\color{red} Shouldn't a vesicle of -1,1,0 have the same activation
+ energy as a vesicle of 0,0,0? It doesn't currently.}
+
+