1. The frequency of the chromosomal types of several western Mediterranean populations of D. subobscura, distributed in a zone running north–south, is analysed and compared with that from an Edinburgh (Knight, 1961) site at a similar meridian as the Mediterranean populations.

2. North–south clines are found in the frequencies of several chromosomal types. Some types are more frequent in the north, decreasing gradually southwards; others show the reverse trend of variation.

3. The comparison with a similar array of populations from Central Europe and the Central Mediterranean area, indicates that the chromosomal types more frequent in the northern populations are mostly the same, i.e. the standard orders. But chromosomal types with complex inversion orders are the most frequent in southern populations: in some chromosomes, different orders are predominant in Israel, southern Italy and southern Spain.

4. The Pyrenees, acting as an ecological barrier, strongly influence the diversity of the populations north and south of the range. This result and the latitudinal clines support the adaptive significance of the chromosomal polymorphism in D. subobscura.

5. The index of free recombination of the population from Malaga (in the south of Spain) is higher than in the populations from the northern Mediterranean area. D. subobscura apparently supports the claims of da Cunha & Dobzhansky (1954) and Carson (1955) that a higher level of chromosomal polymorphism occurs in the central areas of the distribution of a species.

Grizzled (gr) is a recessive gene with complete penetrance but reduced viability. It is linked with waltzer (v) in linkage group X, the recombination frequency being 14·6 ± 2·6%. Its visible effect is to dilute the yellow pigment of the hairs. Body weight is reduced by about 25% atbirth and remains at about 10–20% below that of non-grizzled litter mates. Viability up to the time of classification is about 50–60% of that of non-grizzled litter mates. The mortality of grizzled homo-zygotes occurs at all stages from about 10 days of gestation up to the time of classification. No specific cause of the mortality was evident.

In Neurospora crassa, a change of 2·5°C. in incubation temperature produced highly significant changes in the second division segregation frequency of asco (Linkage Group VI). A minimum value was found at 25°C. consistently increasing towards higher and lower temperatures over the range 15° to 30°C.

Where recombination frequencies in proximal and distal marked regions in Linkage Groups I and VI showed a change, this followed the same pattern as for asco with proximal intervals usually affected; but in one cross a distal region, instead of the proximal, showed a significant change. This behaviour supports the suggestion that heritable factors influencing recombination frequency in specific regions of each chromosome may show differences in temperature sensitivity.

Temperature treatments during protoperithecial development prior to meiosis as well as during meiosis had significant effects on recombination frequencies in several cases. Protoperithecia incubated at one temperature and then transferred to a different temperature after conidiation, gave second division segregation frequencies of asco intermediate in value between those obtained with continuous incubation at either of the two temperatures alone.

In the present work with Neurospora, the similarity in the relationship between temperature and recombination frequency with the reported results in Drosophila is discussed and suggests that the common temperature/recombination frequency curve is indicative of a possible direct effect of temperature on the recombination processes. Present knowledge of these processes is insufficient to permit any detailed explanation of the temperature effects found but the observed negative relationship between recombination frequency and temperature over the range 15° to 22·5°C. is discussed in kinetic terms.

The second division segregation frequency of asco was found to be increased at all temperatures tested when the asci were mounted for scoring in strong sucrose solution compared to mounting in water.

A cytological study has been made, in Aspergillus nidulans, of haploid and diploid homokaryons and of heterokaryons formed between them. The effective region of growth was shown to extend about 600 μ behind the growing edge of the colony and to correspond largely with the tip cells. Acridine fluorescence gave unselective staining of all nuclei; projected nuclear areas, and their distributions in individual hyphae, were determined for the homokaryons and heterokaryons. Diploid and haploid nuclei differ in mean projected area but the overlap of their size distributions precluded unequivocal classification of individual nuclei in diploid–haploid heterokaryons. However, it was shown that the hyphae of a heterokaryon differ in nuclear ratios. This, and other evidence, strongly support the suggestion that heterokaryon maintenance and adaptation occur through interhyphal rather than intrahyphal selection.

The spontaneous reversion to white of three purple mutants of Coprinus lagopus has been demonstrated to be due to suppression. Forty-one of the fifty suppressed purples isolated were crossed to wild-type and all crosses gave fewer than 25% purple progeny, i.e. all the suppressors appeared to be linked to the gene suppressed. Values for the percentage of purples segregated from these crosses varied continuously from 0% to nearly 20%. The functional relationships of the fifty suppressors were investigated by the technique of complementation in dicaryons. The pattern of results formally indicate two suppressor loci with extensive inter-allelic complementation at one of them. The functional allelism indicated by the complementation results is not compatible, in conventional terms, with the apparent positional spread of the suppressors. Several possible explanations are discussed but more data are required before a choice can be made between them. It seems most probable that the same genetic entity is involved in each independent event of suppression and that the characteristic segregations of purples from crosses of suppressed purples with wild-type will have to be explained in terms other than of normal recombination.

A number of acridines have been tested for ability to inhibit conidia of strains of Aspergillus nidulans. The effectiveness of any one acridine in growth inhibition and killing involves interaction of genotype and conditions of treatment such as temperature, pH, treatment medium and light intensity. Mutant alleles which confer growth resistance to acriflavine are selective in their actions towards other acridines, may differ in their dominance relationships with different acridines and are even selective with regard to the conditions under which they confer acriflavine resistance. Certain pairs of acridines, used simultaneously, show additive effects, potentiation, or annulment by one of inhibition caused by the other.

Some of these findings have been applied in a study of factors affecting acridine-induced mutation in Aspergillus conidia. Under conditions which permit metabolism, acriflavine induces a high frequency of unstable morphological variants. One such variant has been shown to be a disomic. Using a system of reversion from auxotrophy to prototrophy, acriflavine-induced mutation has been obtained both with high light intensities and in the absence of light. In the latter case recombination as a feature of the mutation process is excluded.

Strains with altered properties of host-induced modification (HIM) have been isolated. Someof these variants were stable on first isolation but the majority of them are unstable. They segregate during many generations two different stable HIM phenotypes and a third type which continues to segregate. Usually this instability disappears during repeated serial single colonyisolation, but in one case it still persisted after nine such single colony isolation steps. Evidence suggesting that they are unstable partial diploids is discussed.

A technique based upon auxotrophic mutants and recessive suppressor genes has been used to select for somatic recombinants in the dikaryon of Coprinus lagopus. Both haploid and disomic nuclei with a reassortment of chromosomes from the two component nuclei of the dikaryon were found. The presence of the disomics indicates that the recombination process is one of fusion of nuclei with a gradual haploidization by loss of chromosomes at mitotic divisions. An extensive test for crossing-over in the A chromosome failed to give an unambiguous crossover recombinant, but it produced a stock with a normal A chromosome and a fragment of the chromosome.

1. The breeding behaviour of the three tetraploid Rubus species R. caesius, R. calvatus and R. laciniatus (2n = 28) has been investigated.

2. Megaspore mother cells of all three species always undergo a normal meiosis and embryo-sac formation is of the Polygonum type. Egg cells have fourteen chromosomes.

3. There is no evidence for the production of either aposporic or diplosporic embryo-sacs as has previously been assumed.

4. The proportion of sexual and apomictic progeny differs in the three species and, in R. laciniatus, varies according to the chromosome number of the pollinating parent.

5. The apomictic progeny are produced by diploidization of the reduced egg cells. These diploidized egg cells can subsequently be fertilized in R. laciniatus.

6. The versatility in the breeding behaviour of these species is explained on the basis that only one type of embryo-sac is formed and that the developmental behaviour of the egg cell is conditioned by the chromosome number of the pollinating parent. Apomixis in these species is not a consequence of a breakdown of meiosis.

1. Six initially identical populations of Drosophila pseudoobscura have been maintained in population cages for 7 years. Two populations have been kept at 16°C, two at 25°C, and two at 27°C.

2. One and a half years after the start, there was no significant genetic divergence in body size among the populations. When the populations were about 6 years old, a striking genetic divergence in body size was found. The genetic difference between the populations having the smallest and the largest mean sizes is over half the total phenotypic change in size between the two extreme temperatures at which the populations were kept. The populations kept at the lower temperature have genetically larger flies than the populations kept at the higher temperatures.

3. Accompanying the changes in body size were changes in the time of develop ment from egg to adult, the faster developers being the larger flies.

4. The F1 hybrids from crossses between Vetukhiv's populations showed non-additivity of the genes for body size, the F1's in most cases being significantly larger than the midparent. There was no change in variability of body size in the F1 or F2 hybrids.

5. The temperature-directed selection for body size found in Vetukhiv's experimental populations may well be similar in kind to that which has produced temperature-oriented geographic gradients for body size in natural populations of several species of Drosophila.

When mating pairs of Escherichia coli K12 are separated by physical means, e.g. by violent agitation, the Hfr cell is unable to mate with a second F− cell. This effect can be simulated by agitation of the Hfr cells prior to mating and, in addition, such cells can be shown to have an impaired ability to adsorb the male-specific coliphage μ2. Inhibitory concentrations of chloramphenicol block the repair of the damage.

Growth of K(P1) bacteria under conditions which lead to a reduction in the level of nucleases also leads to a reduction of their ability to restrict the growth of λ.C. Experiments designed to estimate the time after adsorption at which restriction takes place indicate that phage DNA is probably restricted by a nuclease while passing through the periplasm.

The restriction of phage λ.C by K(P1) cells is reduced when the cells are subjected to an EDTA cold-wash treatment which has been shown to remove surface-localized enzymes. We conclude that a surface-localized enzyme plays an essential role in host-controlled restriction.

Six one-step Chloramphenicol (Cm)-resistant mutants of Escherichia coli K12 were graded for resistance to Cm, Tetracycline (Tc) and Puromyein (Pm) by streaking on minimal agar plates containing antibiotic. They fell into at least three distinct groups on the basis of their resistance patterns. One mutant showed increased sensitivity to Pm. Most of the mutants expressed their effect on resistance to Cm and Tc in the presence of R-factors carrying resistance genes for these antibiotics, but one mutant with a relatively high level of resistance to Cm had its resistance effect completely masked in the presence of R-mediated resistance. Similar cases were found among mutants selected for Cm-resistance in another strain of K12.