Selection of tests for identifying yeasts.

IN ecological studies, it is often necessary to identify very many organisms. Loader's new work on yeast classification1, 2 recommends about sixty physiological tests for identification, as well as microscopical observations, thus providing a formidable task for anyone bent on identifying hundreds or thousands of yeasts. Hence, to identify a group of yeasts, one must usually select certain tests to use and not others. Here a successful application is described of the systematic method proposed earlier3 for selecting tests to identify strains found in a recent survey of yeasts on strawberries4; we also discuss the relative efficiencies of the tests in Lodder's work, the uncorrected proofs of which were kindly made available for study.