The suitability of stomata length as a criterion in the distinction between diploid and tetraploid rye-grass plants was tested.From the data it appears that diploid and tetraploid plants can be separated with a large degree of certainty if the selection is based on the stomata length .

An autotetraploid population of caraway (Carurn carvi L.) was produced by treating the growing points of young plants with a 0.2% colchicine solution. Selection on morphology, pollen size and pollen shape, followed by a selection on seed size proved to be an effective way to identify the mixoploid C, plants. Three percent of the treated plants produced a tetraploid progeny. Moderate selection in the C, and the C, population resulted in an improvement of the average seed-set to the amount of 50% of the diploid control, with a range of 21-78%. The IOOO-kernel weight of the tetraploid material was 7 g, compared to 3 g in the diploid control. The oil content of the autotetraploid seed was increased with 35.6%; the increase in carvone amounted to 6.9%. The results of the experiment indicate that breeding work in autotetraploid caraway may result in a new cultivar.1NTRODUCTlON

To obtain a Cl-population with a high percentage of tetraploids after colchicine treatment of turnips, a selection in the C o can be used. A comparison is made between two criteria for this selection, viz. the average pollen grain size and the variation in pollen grain size.C0-selection on the basis of the average pollen grain size of five pollen grains per plant reduced the number of diploids in the C1 considerably. When the variation in pollen grain size of these five pollen grains is included as a selection-criterion the number of tetraploids in the C~ increased.The two methods are discussed in the light of the working efficiency and the genetical variation in the tetraploid C~-population. It is concluded that the character of the initial material should govern the choice of the selection criterion.