The phylogenetic relationships of the eight species of the Drosophila melanogaster subgroup are examined on the basis of genetic variation at 33 putative enzyme loci. Values of Nei's genetic distance (ds) range from 0·28 to 1·74. D. sechellia appears closer to D. simulans than to D. mauritiana, the two former being the most closely related. D. orena is quite distantly related to D. erecta (ds = 1). Genetic differentiation supports the existence of three main lineages within the melanogaster subgroup and the yakuba-teissieri pair appears to be closer to the melanogaster lineage than to the erecta-orena one. Inferences of the times of species divergence from allozyme data are made and their agreement to other estimates is discussed.

A new type of genetic control of gene conversion is described from the Pasadena strains of the fungus Ascobolus immersus. It is characterized by cis/trans position effects and incomplete dominance. The P, K and 91 factors segregated from each other like Mendelian alleles and controlled the conversion frequencies and patterns of four nearby, closely-linked white ascospore colour mutations, although they did not usually coconvert with these w sites. Mutations of different origin responded similarly to the same control factors.

These control factors greatly affected the total conversion frequencies, the relative frequencies of the different detected conversion classes and various other conversion parameters. The detailed results are consistent either with P, K and 91 affecting both the frequency of hybrid DNA formation and the correction processes for removing mispaired bases, or if they do not affect the correction processes directly, then they must have large effects on the frequency of asymmetrical hybrid DNA formation, which must usually be much more common than symmetric hybrid DNA, and there must be both an inequality in the frequency with which the two homologous chromosomes (in these crosses, + bearing and w bearing) invade each other, and in the frequency with which the two strands of each chromatid invade the homologue in asymmetric hybrid DNA formation.

Xasta flies appear to segregate five types of gametes in unequal numbers, namely two which contain both or neither of the chromosomes affected by the translocation and inversions, two further classes which contain one affected and one unaffected chromosome, and finally the remainder which have an unbalanced chromosomal content. These conditions are necessary to fit the results observed in crosses involving the Xasta stock. Xasta exhibits balanced polymorphism under crowded conditions and this may be due to the production of toxic substances by Xa larvae which delay the development of wild type-larvae.

The resistance levels conferred by the T-determinants in four R-factors to Tetracycline and Minocycline in cells of Escherichia coli K 12, before and after induction of maximum resistance by treatment with sub-inhibitory concentrations of the drugs, are measured by simple growth-and-challenge tests. The effect of a plasmid TK which confers tetracycline resistance on its host Klebsiella aerogenes is tested in the same way. The five T-determinants fall into a high-level and a low-level group for resistance, the former giving 3- to 4-fold higher resistance in both induced and uninduced cells than the latter. The T-determinants all confer much lower resistance to Minocycline (a tetracycline molecule modified at the C-6 and C-7 positions) than to Tetracycline. The main cause of this difference is that cells carrying a T-determinant exclude Minocycline much less efficiently than Tetracycline, but in addition Minocycline is less effective than Tetracycline in inducing increased resistance. These results are discussed in the light of a model put forward to explain the inducible nature of R-factor resistance to the tetracyclines.

Wild-type chromosomes of D. melanogaster mutagenized by passage through a single generation of hybrid dysgenesis have been compared against identical chromosomes passed through a reciprocal, non-dysgenic cross. Fitness of the chromosome in homozygous condition has been examined in population cages using the technique of balancer chromosome equilibration. The results indicate that amongst chromosomes with no lethal or visible mutation, more than 50% have suffered a measurable decline in fitness. The magnitude of this decline is estimated to be in the range 10–20%.

The relationship between chromosome pairing and chromosome fragmentation has been studied in a gene controlled mutant of pearl millet (2n = 14). Premeiotic mitosis, premeiotic cell development and early prophase I are normal without any fragments, which first appear at pachytene. The extent of fragmentation varies from zero to very extreme with two discrete classes of plants, namely those with partial fragmentation and those with multiple fragmentation. A quantitative analysis of bivalent distribution and the distribution of AI bridges in desynaptic and fragmented cells show all of them to be nonrandom events. We suggest that in cells showing partial fragmentation the bridges and fragments result from U-type exchanges at pachytene. The reduced frequency of AII bridges indicates relatively low sister chromatid reunion at pachytene. In multiple fragmented plants numerous minute fragments were seen from pachytene. Despite these anomalies most PMCs complete meiosis but subsequently abort at the pollen grain stage. The mutant gene also causes disturbances in the sequence of meiotic development in the ear and in the synchronous development of PMCs within an anther. It has no effect on the tapetum or on the physiological development of the anther.

A population cage experiment has been carried out to estimate fitness for a sample of fourteen non-lethal third chromosomes in D. melanogaster. This measurement, which should take into account all aspects of fitness, gives an estimated mean fitness of chromosome homozygotes of approximately ten percent.

Brassica juncea cv. Pusa Bold carrying B. oxyrrhina cytoplasm (oxy cytoplasm) was male sterile and chlorotic under field conditions at low temperature (Prakash & Chopra, 1990). Leaf protoplasts of the chlorotic male sterile alloplasmic line (2n = 36) were fused with hypocotyl protoplasts of green male fertile, B. juncea cv. RLM-198 (2n = 36) using polyethylene glycol. Of the 1043 plants regenerated from 10 fusion experiments, 123 had ‘gigas’ features and were identified as presumptive fusion products. Among field-grown population, one plant was dark green even at low temperatures and male sterile. It possessed 72 chromosomes which formed 36 bivalents at late diakinesis of meiosis-I. This plant was back-crossed to B. juncea cv. Pusa Bold (the maintainer line) for three successive generations. One male sterile, normal green BC3 progeny plant with 2n = 36 was analyzed for organelle constitution. Probing its total DNA with the mitochondrial gene for cytochrome oxidase subunit I revealed that it possessed mitochondria of B. oxyrrhina. Southern hybridization pattern with the gene for ribulose bisphosphate carboxylase oxygenase-large subunit (rbcL) revealed that the chloroplast genome of the chlorophyll deficiencycorrected plant had characteristics of both B. juncea and B. oxyrrhina. The deficiency correction has been attributed to recombination between chloroplast genomes of the two species.

A non-transmissible plasmid, pAV5, was isolated from a hospital strain of Acinetobacter calcoaceticus, JC17. pAV5 confers resistance to the antibiotics tetracycline and neomycin in the genetically characterized strain of A. calcoaceticus EBF 65/65. Transfer of pAV5 can be mediated by the sex factors pAV1 and R751; transfer is occasionally associated with the segregation of the resistance determinants amongst the transconjugants. Phenotypic dissociation of pAV5 corresponds with the formation of two independent plasmids designated pAV51 and pAV52 mediating resistance to neomycin and tetracycline respectively.

A mechanism for RNA–RNA splicing is proposed. A species of RNA (‘splicer’ RNA) hybridizes to precursor mRNA across the splice point. This hybridization can be with intron or exon sequences or both. The double-stranded RNA structure precisely indicates to the splicing enzymes the exact location for exon ligation.

A model for the control of gene expression is presented. The regulation of synthesis of different splicer RNAs will also control which precursor mRNA molecules are spliced. The removal of intervening sequences from a precursor mRNA molecule could be both a signal for that molecule to be transported to the cytoplasm and a means of allowing gene expression.

1. Data from transduction experiments with Salmonella typhimurium can be interpreted in terms of frequencies of recombination by accepting the arguments of Ozeki that all transduced fragments that include a particular locus are equal in length.

2. From a two-point transduction the ratio of the two classes of transductants is not a measure of the ratio of the frequencies of recombination in the two relevant regions but only provides maximum and minimum values for this ratio.

3. From three-point transductions the frequencies of recombination in the regions between the markers, and between the terminal markers and the ends of the fragment, can be estimated.

4. The difficulties of interpreting data from experiments in which a single factor is transduced are briefly discussed.

The genetic relationships of the available second chromosome melanotic tumour mutants in Drosophila melanogaster have been investigated. Complementation tests demonstrate the existence of new alleles of the tu bw locus and show that tu-W and tu-g are alleles. The data suggest that there is a minimum of three major gene loci on the second chromosome involved in tumorigenesis. A number of modifier genes were found which affect the penetrance of the major tumour genes analysed. These and the problems they cause in mapping the low penetrant tumour genes are discussed. It has not been possible to map tu-48a, tu- W and tu-g accurately, due largely to the presence of modifier genes. It appears that the genetic basis for melanotic tumour formation is complex.

A conflict of interest may arise between intra-cellular genomes and their host cell. The example explicitly investigated is that of a ‘selfish’ mitochondrion which increases its own rate of replication at the cost of reduced metabolic activity which is deleterious to the host cell. The results apply to deleterious cytoplasmic agents in general, such as intracellular parasites. Numerical simulation suggests that selfish mitochondria are able to invade an isogamous sexual population and are capable of reducing its fitness to below “5 % of that prior to their invasion. Their spread is enhanced by decreasing the number of mitotic divisions between meioses, and this may constitute a significant constraint on the evolution of lifecycles. The presence of such deleterious cytoplasmic agents favours a nuclear mutation whose expression prevents cytoplasm from the other gamete entering the zygote at fertilization, resulting in uniparental inheritance of cytoplasm. Such a mutation appears physiologically plausible and can increase in frequency despite its deleterious effect in halving the amount of cytoplasm in the zygote. It is suggested that these were the conditions under which anisogamy evolved. These results have implications for the evolution of sexual reproduction. Standard theory suggests there is no immediate cost of sex, a twofold cost being incurred later as anisogamy evolves. The analysis described here predicts a large, rapid reduction in fitness associated with isogamous sexual reproduction, due to the spread of deleterious cytoplasmic agents with fitness only subsequently rising to a maximum twofold cost as uniparental inheritance of cytoplasm and anisogamy evolve.

In the analysis of recombination in Aspergillus nidulans coincidence values of about 1 are found in 3-point tests using markers more than a few map units apart. In comparable tests in which the marked intervals were very short (0·1 map unit or less), coincidence values of over 100 had been found. To account for this difference it was proposed that a necessary condition for recombination, termed ‘effective pairing’, was realized at any particular point on the chromosome in only a small fraction of a population of cells at meiosis. It was supposed that when effective pairing occurred it extended over a very short segment of the chromosome and that the probability of recombination in the effectively paired segment was high, i.e. about 1. Coincidence values greater than unity would be a necessary consequence of such a situation provided the intervals in question had a total length not much greater than that of effective pairing segments.

The experiments described in this paper were undertaken in an attempt to measure the mean length of effectively paired segments, their distribution, and the frequency of exchange within them. The data suggest a mean length of about 0·4 map unit, a mean exchange frequency of about 0·6, and a distribution which is variable, perhaps random.

The occurrence of localized negative interference suggests a way in which a number of difficulties encountered in relating the experimental evidence concerning the time and mechanism of recombination with the cytological evidence concerning the sequence of events at meiosis might be resolved. The data indicate that the frequency of recombination between linked loci is a measure principally of the frequency of effective pairing between them. If effective pairing is synonymous with homologous contact between chromosomes, and evidence is presented which suggests this may be the case, it becomes possible to construct a simple model which is compatible with the view that recombination takes place before chromosomes are paired, in the cytologically observable sense (i.e. before zygotene), at meiosis.

The recombination events occurring within effectively paired regions are generally, although possibly not exclusively, reciprocal. Non-reciprocal recombinants have been encountered in Aspergillus and other organisms, characterized by the occurrence of 3:1 ratios in tetrads. On the basis of evidence currently available it does not seem necessary to invoke a special mechanism of recombination, distinct from crossing over, to account for the formation of non-reciprocal recombinaiits. A single mechanism of recombination of the copy-choice type which, although primarily a reciprocal process, is nevertheless not necessarily exactly so or always so in detail, will account for the observed results.