The genetic basis of variation in rate of seedling growth, and development has been examined in the Australian commercial population of Phalaris tuberosa. A model of additive genetic maternal effects has been used, with seed weight of the female parent as an index of maternal ability. Rate of leaf appearance, rate of tillering and growth per tiller are all genetically variable in the population, with estimated heritabilities of 0·36, 0·23 and 0·34 respectively on an individual seedling basis. Total seedling growth has a lower heritability (0·17), due to a negative genetic correlation between tiller production and growth per tiller ( − 0·46). These two components have also been shown to be subject to qualitatively different seed size maternal effects. Genetic differences in seed size in the female parent have been found to influence growth per tiller, while environmental differences in seed size affect primarily the rates of leaf appearance and tiller production.

Recombination data from crosses made at a single constant temperature of incubation were compared with those from crosses transferred to a different temperature at either the time of conidiation of protoperithecia by the strain of opposite mating-type, or after fertilization when crozier stages were first visible. Results were also compared from reciprocal crosses, from crosses made in different ways and from crosses in which protoperithecia were conidiated at different stages of maturity.

Different temperature regimes during vegetative growth and proto-perithecial development had highly significant effects on subsequent meiotic recombination, while temperature differences during later premeiotic stages (between conidiation of protoperithecia and the crozier stage) had no or little effect. It was found that premeiotic controls could have as great, or greater, effects on meiotic recombination than those operating directly during meiosis. The possible adaptive significance of this is discussed.

Recombination frequencies were affected by the method of making a cross (joint-inoculation of strains of opposite mating-type, or conidiation of protoperithecia), and by protoperithecial age at the time of conidiation by the opposite mating-type. Differences in recombination between reciprocal crosses were obtained and were dependent on temperature of incubation and age of protoperithecia at the time of conidiation. Recombination was not affected by different lysine concentrations in the medium. Genetic differences in premeiotic effector-production between the strains used were inferred.

Genetic systems involving developmental inactivation of entire chromosomes occur in two widely different groups of organisms: mammals and coccids (Homoptera: Insecta). The two groups show several similarities and some interesting contrasts with respect to this unusual cytogenetic phenomenon. Although mammalian X chromosomes and coccid paternal sets are components of different genetic systems, comparisons between them nevertheless suggest approaches that might prove to be of value. Further, the occurrence of facultative heterochromatization in these two wholly unrelated taxa must mean that this type of heterochromatization represents a fundamental capacity of chromosomes.

Associative overdominance due to linked detrimental mutations was investigated using the method of moment equations based on diffusion models. The expectation of the apparent selective value at the marker (neutral) locus has been evaluated. Assume two linked loci, at one of which the steady flux equilibrium is reached under constant mutational input of deleterious mutations (with rate v) having disadvantages hs in heterozygote and s in homozygotes. At another locus, the neutral alleles are segregating with frequencies near 0·5. Let Ne be the effective size of the population and c be the recombination fraction between the two loci. Then the coefficient of associative overdominance at the neutral locus can be obtained by taking the expectation with respect to chromosome frequencies at steady flux equilibrium. It becomes approximately

where (LI−L0) is the inbreeding depression caused by deleterious mutations under complete inbreeding, and Nehs ≫ l and hs ≫ v are assumed. More generally, if the inbreeding depression of a chromosome segment with a length of recombination fraction C is (LI−L0) then s′ at the neutral marker at the edge of the segment is

where hs is the average heterozygote disadvantage of detrimentals.

The significance of the associative overdominance is discussed in relation to actual observations. It is proposed that the most of the observed heterozygote superiority including inversion chromosomes of Drosophila, isozyme alleles in Avena and ABO blood group genes in man could be explained by the associated detrimentals.

Recombination between chloramphenicol-sensitive (Cms) mutants of Rl, and R100, has been demonstrated in Escherichia coli K12 rec+; it occurs at reduced frequency in recB and recC, and is not detectable in reeA, indicating that R factor recombination depends on host functions. Some mutants of R1 also recombine with an R100 mutant in a similar way. recA cells carrying an R1 and an R100 Cms mutant (hetero-R state) have a low level of chloramphenicol-resistance, and form a chloramphenicol acetyl transferase that has lower specific activity than enzyme from hosts carrying wild-type or recombinant factors. These results suggest the occurrence of interallelic complementation between mutant R factors.

Haploid segregante from vegetative diploids of Ustilago violacea appear as spherical colonies (papillae) growing above a background of dead diploid cells on complete medium containing DL-p–fluorophenylalanine (PFP). Most mutants segregate normally but one-third of the mutants were expressed infrequently in 0–5% papillae only. These mutants, designated ‘missing-markers’, were found to be on either of two chromosomes that remained disomic after treatment with PFP. When cells from a disomic papillum were streaked on complete medium, monosomics in which ‘missing-markers’ were expressed segregated spontaneously at a low frequency. Thus, of 10–12 linkage groups identified in U. violacea, two remain disomic after PFP treatment. Possible reasons for these differences between chromosomes in the same genome are discussed.

Haploid and diploid stock cultures did not differ either in resistance to PFP, or in the production of papillae on PFP medium. Haploid segregante from a diploid were slightly more resistant to PFP than the wild-type haploid cultures under some conditions, but were very different in that they no longer produced papillae on PFP medium. These haploid segre-gants resembled one of three PFP-resistant mutants (pfp-A) isolated from a wild-type haploid stock grown on PFP medium. The significance of these results to the mechanism of haploidization in U. violacea is discussed.

Plants of Triticum aestivum (2n = 6x = 42) ditelocentric 5BL were treated with EMS in order to produce mutations in the 5B system by which meiotic pairing between homoeologous chromosomes is normally prevented. To check for the occurrence of mutation T. aestivum ditelo-5BL plants were pollinated with rye (Secale cereale 2n = 14) and meiosis was examined in the resulting hybrids.

Wheat-rye hybrids were scored for the presence of mutants when the wheat parents were either the EMS-treated wheat plants, or their selfed derivatives, or their progenies obtained after pollination with untreated euploid individuals.

Mutants were detected by each of these procedures and mutant gametes were produced by the treated ditelocentric plants with frequencies between 1·5 and 2·5%, but there were differences between the mutants in the extent to which homoeologous pairing occurred in the derived wheat-rye hybrids. The differences may have resulted from the occurrence of mutation at different loci or to different extents at the same locus.

Two mutants, Mutant 10/13 and Mutant 61, were fixed in the homozygous condition. Mutant 10/13 was made homozygous both in the 5BL ditelocentric and in the euploid conditions but these genotypes regularly formed 21 bivalents at meiosis, and there was no indication of homoeologous pairing although the mutant 10/13 gave rise to homoeologous pairing in wheat-rye hybrids.

An investigation was made of the chromosomal position of the mutant locus, in Mutant 10/13 of Triticum aestivum (2n = 6x = 42), affecting homoeologous chromosome pairing at meiosis. In hybrids between Mutant 10/13 and rye (Secale cereale 2n = 14), homoeologous chromosomes frequently pair at meiosis although normally, in wheat-rye hybrids, this happens infrequently.

The association of the mutant condition with chromosome 5B was determined by (i) the absence of segregation in hybrids obtained when Mutant 10/13 monosomic 5B was pollinated by rye; (ii) the occurrence of trisomie segregation for pairing behaviour in 28-chromosome wheat-rye hybrids, obtained from SB trisomie wheat parents with two 5B chromosome from a non-mutant and one from a mutant parent; (iii) the absence of segregation for pairing behaviour in the 29-chromosome wheat-rye hybrids obtained from the same trisomie wheat parents.

The alternative pairing behaviours segregated independently of the centromere when wheat plants that were simultaneously heteromorphic, 5BL telocentric/5B complete, and heterozygous for the Mutant 10/13 state, were pollinated by rye. The alternative chromosome-pairing patterns segregated to give a ratio not different from 1:1, so that the association of homoeologous pairing with Mutant 10/13 probably derived from the occurrence of mutation at a single locus on 5BL. In the disomic heteromorphic state, 5BL was 91 map units in length.

Trisomie wheats with two complete 5B chromosomes and one 5BL telocentric, that were also heterozygous for the Mutant 10/13 condition, were pollinated by rye. Among the resulting 28-chromosome hybrids there was a 2:1 segregation of hybrids with low pairing: high (homoeologous) pairing and also of hybrids with complete 5B: telocentric 5BL. However, there was no evidence of linkage in this trisomie segregation. All the 29-chromosome hybrids from this cross had low pairing and it could be concluded that the single mutant allele, in Mutant 10/13, was recessive. In the trisomie condition, relative to a simplex situation, 5BL was 33·05 map units in length.

The critical locus on 5BL was designated Pairing homoeologous. The normal dominant allele was symbolized Ph and the recessive allele, in Mutant 10/13, ph.

The prevention of homoeologous pairing by the activity of a single locus makes the evolution of the regular meiotic behaviour of T. aestivum more readily comprehensible.

The glucose analogue 2-deoxy-D-glucose seriously inhibits the growth of Coprimts lagopus. Following treatment with N-methyl-N′-nitro-N-nitrosoguanidine 388 resistant mutants were isolated. It is shown that the mutants isolated are probably allelic; they were phenotypically similar and no complementation was observed. The mutants were pleiotropic in the sense that although they were initially selected only for resistance to 2-deoxy-D-glucose they were found to be cross-resistant to both of the related analogues, sorbose and glucosamine. Furthermore, the mutants were unable to utilize fructose as a sole carbon source. It is demonstrated that the inability to utilize fructose results from a defect in sugar transport. The gene symbol ftr is proposed for this cistron and it is shown that though the gene is quite closely linked to its own centromere, it is unlinked to centromere markers of the six known linkage groups.

Chain-termination suppressors which are almost certainly amber suppressors have been isolated in Salmonella anatum by a technique involving the use of amber mutants of bacteriophage 01. The same technique could be used in any species of Salmonella (and many strains of Arizona) and analogous techniques are suggested for use in other genera of bacteria and in higher plants.