The question of what is meant by random fluctuations in selection intensities in a finite population is re-examined. The model presented describes the change in the frequency of a gene in a haploid population of size M. It is assumed that in any generation the adaptive values of A and a are equally likely to be 1 + s: 1 or 1: 1 + s. If s is the selective advantage and x the frequency of gene A, then the first two moments of the change in frequency are found to be m(Δx) = x(1 − x)(1 − 2x) θ/2M and

where E(s2) = θ/M. The ultimate probability of fixation is computed, showing that variability in selection increases the chance of fixation of a rare gene. A more general form for m(Δx) also is obtained. This form is compared with the equation currently used in describing random fluctuations in selection intensities.

Initial Fertility (IF) strains of Streptomyces coelicolor are able to convert recipient strains (UF) to the IF condition by contact, without concomitant transfer of chromosomal markers. The conversion is prevented by the presence of acridine orange in the medium of the mixed culture. Acridine orange is also moderately effective in inducing the formation of UF variants from IF-treated strains. No effect of the drug is observed on UF variant formation from Normal Fertility (NF) strains nor on the behaviour of the fertility factor in NF × UF mixed cultures. The hypothesis is put forward that the fertility factor works as an episome in S. coelicolor, fixed to the chromosome in the NF strains, free in the IF strains and missing in the UF strains.

Forty-one suppressors obtained after NG and EMS mutagenesis of two ‘polar’ mutants of segment ‘29’ in Podospora anserina were genetically analysed. Three classes of suppressor could be distinguished on spectrum pattern criteria. One representative suppressor of each class was demonstrated to be non-gene specific. The class I suppressor was dominant and only suppressed polar mutants in segment ‘29’ and non-ICR-induced ones in genes where polarity cannot be determined. Class II and III suppressors were partially dominant and they suppressed polar, non-polar and even ICR-induced mutants. The difference between classes II and III seems to be only quantitative. According to whether class II and III suppressors are considered strongly or weakly allele-specific, two hypotheses are considered. First, tRNAs could be involved in all three classes of suppression: class I would be nonsense-specific and classes II and III would be nonsense-missense suppressors. Secondly, tRNA could be involved only in class I suppression, while ribosomal ambiguity could be responsible for class II and III suppression.

Interspecific aneuploids were made in the genus Cucurbita by pollinating the autoallotroploid (diploid C. moschata – haploid C. palmata) with pollen from diploid C. moschata. Seventy-eight, or 0·36%, of the potential ovules showed some degree of continued development after pollination. The high frequency of continued ovule and embryo development indicated a preferential distribution or loss of C. palmata univalents during meiosis in the autoallotriploid and that megagametes aneuploid for more than one C. palmata chromosome were capable of initiating continued ovule development after fertilization. Lethal effects of interspecific aneuploidy were expressed by complete embryo abortion to sterility of mature plants. Nine plants developed to maturity and seven were cytologically identified. Four were trisomic and two were monosomic for single C. palmata chromosomes, i.e. they were 2n + 1 and 2n + 19 (n = 20) respectively. One plant was 2n and phenotypically identical to the C. moschata parent. One 2n + 1 plant was fertile and the single C. palmata chromosome was transferred in the succeeding generation.

In this paper an exact treatment is given for the stochastic behaviour of the frequency of haploid genotypes in an infinite population when the absolute fitnesses of the two genotypes vary at random over generations. The main qualitative result from this treatment is that natural selection will favour that allele with the largest geometric mean fitness. A diffusion equation is derived whose solution is identical to the exact solution. The drift coefficient for this equation is of the form − μp(1 − p) + σ2(½ − p)p(l − p). This differs from the drift coefficient used in previous treatments of this problem and reduces the rate of quasi-fixation. Various waiting time problems are solved using this diffusion equation.

Spontaneous reversions from instability to stability were studied in a mutant homothallic strain of yeast. The mutant strain was characterized by persistent lethal sectoring attributable to the presence of a recessive locus uns1. It was found that (1) the meiotic reversion rate is about 140 times higher than the mitotic reversion rate, (2) reversions are much more common in the homozygote uns1 uns1 than in the heterozygote UNS1 unsl, (3) revertant segregants and lethal segregants tend to occur together, and (4) meiotic revertants tend to occur as twin revertants. These results indicate that reversion from unstable to stable results from genetic recombination and that, in meiosis, this recombination involves an unequal exchange at the first division.

Penetrance and expressivity of the Drosophila mutant Abnormal-abdomen (A53g) may be reduced by adding inhibitors of protein synthesis, of RNA synthesis, and of oxidative phosphorylation to the defined, sterile medium on which larvae are developing. When grown on regular diet, the mutant flies contain a higher concentration of total protein than do wild-type flies. The metabolic inhibitors which reduce the expressivity of the mutation also lower protein concentration in these mutant flies. The reduction of protein synthesis is directly correlated with the reduction of expression of the mutant genotype. The regulation of the morphological abdominal abnormalities by the complex A53g genotype is discussed in relation to the control, by this same genetic system, of the mechanisms of protein synthesis.

The shape of the mandible in. nine sublines of C57BL/Gr, seven other strains of ‘C57 ancestry’ and four unrelated strains was studied by multivariate techniques. The generalized distance function was used to classify individuals in the groups which they most closely resembled. The degree of misclassification depended on the pedigree relationship between strains and sublines. The generalized distance between pairs of subline centeroids was also highly correlated (r = 0·60) with the number of generations between them. A canonical variate analysis was used to reduce the dimensionality so that a graphical display of the relationships between strains and sublines could be made. The results agreed closely with the classification analysis. It was concluded that the shape of the mandible could be used for subline identification though the accuracy of this technique depends on how closely the sublines are related.

Simultaneous selection by independent culling levels of two correlated traits in all four combinations of directions was investigated with Tribolium in a replicated experiment extending over nine generations. In addition to the two primary traits, 13-day larval weight and pupal weight, four secondary traits (pupation time, adult emergence time, adult weight and larval number) were observed.

The observed responses for both selected and unselected traits agreed with theoretical expectations after the latter were adjusted for changes which occurred in genetic and phenotypic parameters. Phenotypio variances for the selected traits were correlated positively with population means, yet genetic variances and heritabilities declined in all selected populations. No change was detected in the genetic correlation between selected traits even though the divergent two-trait selection was designed especially to ‘break’ the positive correlation of + 0·55 ± 0·12 present in the base population.

Striking changes in growth and developmental patterns resulting from the divergent selection were discussed in terms of metamorphic limits and ‘stabilizing’ genetic correlations.

The stimulation of in vitro tRNA aminoacylation by post-microsomal supernatant enzymes isolated from Abnormal Abdomen (A53g) adult flies is a function of the residual genome of these flies rather than of the major mutant gene, A53g. Genes controlling this stimulation are located on the X chromosome as well as on the autosomes. These observations are discussed in terms of the phenotypic effect of the major mutant gene in response to changes in protein synthesis which are under the control of modifier genes responsible for aberrant genetic translation mechanisms.

Intergeneric recombination of chromosomal genes between strains of Klebsiella pneumoniae and Enterobacter aerogenes has been observed. In these experiments a male strain of K. pneumoniae acted as donor and E. aerogenes acted as recipient. Closely linked genes were co-transferred with very high frequency, and transfer of a large fragment carrying several separated genes was not uncommon. Subcultures of a hybrid which had received a large fragment carrying five markers were tested for genetic stability; no spontaneous segregants were found among 34000 colonies tested. These results suggest that the transferred fragment had been integrated, and indicate the close genetic relationship between K. pneumoniae and E. aerogenes. This is in agreement with their very similar biochemical properties.

Brassica juncea Coss (2n = 36) is a naturally occurring allotetraploid. In the progenies of F1 hybrids of its constituent parents B. campestris sub.sp. oleifera var. toria × B. nigra and B. campestris sub.sp. japonica × B. nigra, the majority of the plants were parthenogenetic homozygous amphidiploids like B. juncea whereas a few differed in morphological and physiological characteristics. The latter, presumed to have arisen partheno-genetically, were suspected to be homozygous for chromosomes changed by allosyndetic recombination. One of these derived plants with curley leaves was crossed first with one of the homozygous amphidiploids and secondly with B. c. sub.sp. oleifera var. toria. Sporocytes of the two hybrids were studied at meiosis. The presence of one or two quadrivalents in the first hybrid and two trivalents in the second suggested that allosyndetic recombination had occurred in the 18-chromosome hybrid. The progeny was presumed to have arisen through pseudo-diploid parthenogenesis and this in association with allosyndetic recombination may provide a means of breeding higher yielding genotypes.

Doublets of Oxytricha fallax possess two complete sets of ciliature. The doublet phenotype is inherited through sexual and asexual reproduction as a cortically determined trait. The trait is also inherited through cystment, independent of cyst size. Prior work shows an absence of all visible cortical organelles, except cell membranes, in the cyst; thus the visible structures are not themselves determinative.

Results from excision experiments performed on encysting doublets indicate that the determinative difference between doublets and singlets is the presence of one or two ‘determinative regions’ located on the ventral surfaces of the organisms which serve as sites for the initiation of ciliary primordium development. Doublets possess two such areas, but singlets possess only one.

Amongst a collection of temperature sensitive (TS) mutants of Escherichia coli K-12, some have been found which can grow at the restrictive temperature (42 °C) if the osmotic pressure of the medium is raised by the addition of sodium chloride (1%) or sucrose (12·5%). These mutants are described as temperature sensitive osmotic remedial (TSOR) mutants. At the restrictive temperature they are not osmotically fragile, but do display decreased resistance to inhibitory agents such as deoxycholate, actino-mycin D and acridine orange; they also show release of the periplasmic enzyme ribonuclease. These results indicate a change in the cell's outer permeability barrier. The genes affected in six of the mutants have been located on the E. coli linkage map. The mutations, which occur at loci not previously described, have been named envM–envT to indicate their effect on the cell envelope.

An infinite population practising a constant amount of selfing and random mating is studied. The effects of the mating system on two linked loci with an arbitrary number of neutral alleles are determined. Expressions are obtained for the two-locus descent measure, and hence genotypic frequencies and disequilibria functions, in any generation. The nature of the equilibrium population is deduced. The special cases of pure selfing or pure random mating and completely linked or completely unlinked loci are considered separately.

Of 150 wild-type strains of Pseudomonas aeruginosa examined, 48 formed recombinants when mated to P. aeruginosa strain PAO FP− and hence presumably possess sex factors. Three different types of sex factor were distinguished by the pattern of transfer of particular markers in different regions of the chromosome and by the ability to confer resistance to mercury in strain PAO. One new sex factor, FP39, was studied in detail, and while similar to the previously studied FP2 in terms of transfer kinetics, natural stability and resistance to curing by acridines, it differed from FP2 in promoting chromosome transfer from a site 10 min to the left of the FP2 origin and in showing apparently aberrant entry kinetics for a leucine marker situated 48 min from the FP2 origin. This was due to FP39 having a genetic determinant either for a structural gene of leucine biosynthesis or a specific suppressor gene for this locus. PAO strains carrying both FP2 and FP39 were unstable for both sex factors, suggesting a relationship between them.

Cells carrying traJ− mutants of F are transfer-deficient and are good recipients in conjugation (Achtman, Willets & Clark, 1972). In addition, the traJ gene product is involved in pilus- and in plasmid-specificity (Willetts, 1971). J-independent mutants were isolated as revertants of a traJ− mutant; they still carry the traJ− mutation but also carry at least one other mutation which results in transfer in the absence of the traJ gene product. J-independent transfer of these mutants is not inhibited by the R 100 repressor. Various models are presented which can account for the properties of traJ− mutants and of these J-independent revertants.

Male Glossina morsitans were subjected to various doses of gamma radiation in air or nitrogen and mated to untreated females. The sex ratio of the F1 progeny was biased towards males, and this may be explained by the extra vulnerability to dominant lethal induction that the presence of the large X chromosome gives to female determining sperms. The mean fertility of the F1 progeny was subnormal. This was due to the induction of 50 % sterility in a large proportion of F1 individuals. Cytogenetic examination of the progeny of outcrosses of F1 individuals indicated that in most of the semi-steriles there were translocations involving the large autosomes or the Y chromosomes. Almost all the normally fertile F1 flies gave only cytogenetically normal progeny. The F1 sex-ratio distortion and semi-sterility would provide a ‘bonus’ in the application of the sterile male technique, which would amount to a 15–50% saving in the releases required to achieve a population control target compared with the requirement if the F1 was normal.

A derivative line of the homothallic Colonia strain of Physarum polycephalum has been isolated which produces plasmodia with high efficiency within clones of amoebae. Using the synergistic effect of ultraviolet light and caffeine, mutants of this line have been isolated which fail to undergo the developmental transition between haploid amoebae and diploid plasmodia (apt mutants). They are isolated by selecting for amoebae which fail to produce plasmodia within clones. Complementation tests of four mutants have shown that they are mutants of four different loci and they are recessive to wild-type. A further analysis of one mutant reveals that the apt-1 locus is unlinked to three other known markers. Crosses of this mutant with heterothallic strains yield progeny which are homothallic indicating that the lesion is not a revertant from a homothallic to a heterothallic mating-type. The use of this system in isolating developmental mutants is discussed.

An approximate expression, is derived for the rate of change in frequency of an inversion introduced at a low initial frequency into a multi-locus system at equilibrium under recombination and selection. It is shown that this expression gives accurate predictions of the rate of progress of the inversion, even if the initial population is perturbed somewhat from equilibrium. Extensions to the cases where there are sex differences in recombination and selection are considered. An implication of the results is that selection pressure for newly arisen inversions depends on the existence of a stable equilibrium with linkage disequilibrium. The expected chance of survival of a new inversion in a large population is shown to be approximately one half the square root of the loss in fitness due to recombination.