In the December 1968 issue of BioScience, an article appeared entitled "Calculation and Miscalculation of the Allometric Equation as a Model in Biological Data" by Jerrold H. Zar. This article, while dealing with a very basic and important problem of data analysis-the question of whether or not to use data transformations in fitting models to biological data and the effect of these transformations-tends, I believe, to be misleading for the biological scientist. Since this journal is read primarily by reseachers in the biological sciences, it is important that some of the assertions in the referenced article be clarified. The discussion here falls into five categories: 1) Choice of model 2) Choice of estimation procedure 3) Comparison of models 4) Interpretation of the comparisons made in the referenced article 5) Some features of iterative least