In mixture experiments there is a constraint that the variables are the proportions of components that are mixed together with the consequence that these proportions sum to one. When fitting regression models to data from mixture experiments we may be interested in reprenting the fitted model with a surface plot. The constraint on proportions means