Abstract Our understanding of meiosis springs from two suggestions made by Weismann in 1887. One was that meiosis would be found to compensate for fertilization in the life cycles of both sexes and all organisms. The other was that the development of sexual reproduction in evolution depended on the value of meiosis in exposing the results of genetic recombination to natural selection. In confirming these propositions we were bound to discover that the properties of meiosis appear both as the causes and the consequences of evolution: it is the hinge on which turns the evolution of breeding method, reproductive habit, life cycle and hereditary structure, that is the genetic system, in all sexually reproducing species of organism. We have had three main fields of attack on our problem. First, there was the natural variation of meiosis including that of two-track heredity within the species: here, animals took the lead. Secondly, there was the experimental field - both with genetic controls such as polyploidy and the sterilizing mutations of mitosis as well as meiosis, and with physical and chemical controls : here, the higher plants and microorganisms have given us our great opportunities. Thirdly, we have the widening field where physicochemical knowledge and genetic control converge and collaborate. In all this work we have to be aware that meiosis works with chromosomes which always have the two functions of accomplishing evolution and of implementing its results in heredity. In consequence, the adaptation of meiosis is perpetually imperfect.