ON November 24, Prof. G. D. H. Carpenter, Hope professor of zoology in the University of Oxford, delivered the second part of his inaugural lecture (see NATURE of November 25, p. 813). Dealing with the imitation of inedible or distasteful by edible species (Batesian mimicry), and the resemblances between inedible models (Mullerian mimicry or synaposematism), he laid stress on the fact that edibility and the reverse are not absolute but relative. The results of experiment accord well with theory, as is shown by Prof. Carpenter's own observations with monkeys, Moreton-Jones's with birds and Cott's with frogs. The fact of preferential feeding is well established. The polymorphism often shown by mimetic species is also in accordance with theory. It is evident that while models would gain, on the principle laid down by Miiller, by diminishing diversity between their appearances, mimics on the other hand would find advantage by increasing it. Mimicry cannot be simply the result of coincidence due to a limited range of colour, nor can it be accounted for by supposing a parallel drift of variation. Mimicry might deceive an artist, not an anatomist. The only available key to the diverse phenomena of mimicry is the principle of natural selection.