Re-examination of the "3/4-law" of metabolism.

We examine the scaling law B is proportional to M(alpha)which connects organismal resting metabolic rate B with organismal mass M, where alpha is commonly held to be 3/4. Since simple dimensional analysis suggests alpha = 2/3, we consider this to be a null hypothesis testable by empirical studies. We re-analyse data sets for mammals and birds compiled by Heusner, Bennett and Harvey, Bartels, Hemmingsen, Brody, and Kleiber, and find little evidence for rejecting alpha = 2/3 in favor of alpha = 3/4. For mammals, we find a possible breakdown in scaling for larger masses reflected in a systematic increase in alpha. We also review theoretical justifications of alpha = 3/4 based on dimensional analysis, nutrient-supply networks, and four-dimensional biology. We find that present theories for alpha = 3/4 require assumptions that render them unconvincing for rejecting the null hypothesis that alpha = 2/3.