Selection for plants showing low or high transmission rates of B chromosomes demonstrates the existence of genotypes which affect this character. The results suggest that the gene or genes involved in this control are located in the A chromosomes. It is shown that plants from the low transmission rate class tend to lose B chromosomes, while plants from the high transmission rate class tend to accumulate Bs. It is therefore concluded that these genotypes can influence the frequency of B chromosomes in different populations.

B chromosomes of S. cereale have been introduced into S. vavilovii, which lacks them in nature, to compare their effects on fitness. In both species, viability was unaffected, while fertility progressively decreased when the number of B chromosomes increased from two to four. These effects depend on the B chromosome number of the individuals, and not on the frequency of B chromosomes in the population.From the similarity of B chromosome effects on S. cereale and S. vavilovii it is argued that the maintenance of B chromosome polymorphisms mainly depends on B chromosomes themselves, and not on a coadaptation between B chromosomes and the cereale genome.

Twenty 9 OB x d 2B crosses were made in which a single male was crossed with four different females, and the character 'mean number of Bs per plant in the progeny' was analysed. The 2B plants used as males and the OB females belonged either to low or to high transmission rate lines, previously selected. The genetic basis of the difference between the high and low transmission rate lines was statistically tested. The male group (high or low class of the 2B male) has a significant effect on the progeny, while the female group (high or low class of the 013 female) has no significant effect. The genes controlling the transmission rate of B chromosomes affect the proportion of OB vs. 2B plants in the OB X 2B cross, but this is not a result of any effect on the nondisjunction mechanism, since the proportion of lB plants is low in all cases. Structural variants of B chromosomes appeared de novo at a high frequency.

Male meiosis and gametogenesis were studied at metaphase I, metaphase and anaphase of the first pollen grain mitosis, and bicellular and tricellular pollen grain stages in 2B rye plants belonging to the low (L) and high (H) B transmission rate lines previously selected. Our results show that B chromosome behaviour significantly differs in both lines whereas the behaviour of the normal complement does not differ. In the L line the Bs form univalents in 81.07 per cent of the metaphase I cells, and are conserved in 44.14 per cent of the pollen grains at first metaphase whereas the remaining Bs are eliminated as micronuclei. In the H line the Bs form bivalents in 87.71 per cent of the metaphase I cells, and are present in 82.48 per cent of pollen grains at first metaphase. The Bs of the L and H lines do not differ in their ability to undergo nondisjunction at first pollen grain anaphase. This indicates that the different B transmission in the L and H lines results from their differential ability to form uni-or bivalents at metaphase I, which determines their loss or conservation in the pollen grains. The L and H lines also differ in pollen viability at the tricellular stage because 19.75 per cent of pollen grains of the L line and only 1.2 per cent of the H line are inviable.

Four Korean rye populations with B chromosomes at frequencies ranging from 15 to 55 per cent have been analyzed in two successive generations in a field experiment. B chromosome frequencies were maintained in spite of the environmental change from Korea to Madrid. Values of variables with effect on viability and fertility have been estimated in plants with and without Bs, in both generations. B chromosomes did not affect viability, while fertility decreased when the number of Bs increased. Our data are against an adaptive nature of B chromosomes.First pollen mitosis was observed in 2B plants. The frequency of metaphases with lB and the rate of nondisjunction at anaphase were high and similar in all populations. Since the decrease of fertility due to Bs and the cytological behaviour of Bs at gametogenesis were similar in all populations, the balance between these two factors cannot account for the establishment of equilibria at different frequencies.