1. Tetraploid (4n) plants in Datura Stramonium, in comparison with diploid (2n) plants, are characterized, among other morphological differences, by spherical instead of ovate capsules and by larger pollen grains. 2. Pollen of tetraploids is relatively good, containing only slightly more shriveled grains than that of diploids. 3. The chromosomes in the pollen mother cells are typically arranged in twelve sets of four each, and at the reduction division in about 70 per cent of the cases investigated assort 24-24, in about 30 per cent of the cases assort 23-25, and rarely 22-26. 4. Tetraploids have occurred spontaneously at least six times, three of them in three distinct lines and three of them as plants derived from crosses between two other lines. 5. Since one of the tetraploids arose in an F₂ from the cross P₂A₂ x p₂a₂, and itself was shown to be P₂p₂A₄, the origin of tetraploidy in this case cannot be attributed to a doubling of the chromosomes before reduction. Since two other tetraploids arose in the F₁ from the cross P₂a₂ x p₂A₂ and were themselves P₂p₂A₂a$_2$, in these two cases the origin of tetraploidy cannot be attributed to apogamy of a 4n egg cell. Other methods of origin are suggested. 6. In comparison with diploids, tetraploids (whether selfed or crossed with other tetraploids) set a relatively small proportion of capsules, which contain a relatively small number of seeds. 7. Crosses between tetraploids and diploids apparently are impossible when the diploid is the female; when the tetraploid is the female, an average of one viable seed has been obtained from about four pollinations. The offspring from this cross have been diploids, triploids, and (2n+1) mutants. 8. Genetic and phenotypic ratios for quadriplex (A₄), triplex (A₃a), duplex (A₂a₂), simplex (Aa₃), and nulliplex (a₄) parents, selfed and crossed inter se, are derived from the assumption of a random assortment of chromosomes at the reduction division. 9. Tabulations of breeding data in some 37,400 individuals for tetrasomic inheritance of the Mendelian pair purple-white flower color confirm the theoretical ratios, and justify the assumption of random assortment of the chromosomes which carry the genes for purple pigmentation. 10. The inheritance of the pair armed-inermis is in accord with the inheritance of the purple-white pair, aside from the production of about 2 per cent exceptional recessive offspring from A₃a parents when back-crossed to a₄ individuals. A possible explanation is suggested but no experimental evidence on the question is as yet available. 11. The possible relation of tetraploidy to the evolution of plants is discussed, and breeding and cytological tests are given to distinguish tetraploidy from "double diploidy." Union of homologous chromosomes is suggested as a possible cause of gigas-like forms without the double number of chromosomes.

	Year	Citations

References