Contrary to opinions expressed by various authors, the phenotype of heterozygotes for mammalian sex-linked genes gives no support for the Lyon hypothesis (L.H.). Evidence, mainly from the mouse, shows that in such heterozygotes, both alleles act together as in autosomal genes.

In the present paper, it is shown that neither the behaviour of double heterozygotes for sex-linked genes nor that of X-autosome translocations provides independent evidence in favour of the L.H.: in each case, the interpretation depends on that of the behaviour of single heterozygotes and hence fails to discriminate. Moreover, new facts from both types of situation are also contrary to the L.H. In particular, a unified interpretation which fits the behaviour of genes in all known types of X-autosome translocations in the mouse requires the assumption that partial inhibition of gene action happens in both X-chromosomes of mouse females, and presumably the females of other mammals. The new hypothesis is consistent with all relevant genetical facts and, like the L.H., it also accounts for dosage compensation.

1. Two methods are examined of introducing new genetic variance into a line of mice selected for high 6-week weight which, at its limit, displayed no additive genetic variance.

2. The first method—irradiation—gave largely negative results. Any further gain under selection that was achieved could not be clearly distinguished from a possible environmental trend.

3. The second method—outcrossing to an unselected strain and then selecting from the cross—resulted in a clear gain over the original limit, but nine generations were required even to recover the original limit.

4. Various methods of transcending selection limits are evaluated in terms of their application to livestock improvement.

1. Repeated measurements on the growth of individual leaves in seedlings and young plants of ryegrass combined with dissections of the apex of the shoot and of very young leaves have shown that the basic pattern of leaf formation is very similar in Lolium multiflorum and L. perenne.

2. The initial growth rate of a leaf primordium is low but increases suddenly at a point when the primordium is somewhat longer than the apex and about 1 mm. in length. After this transition, which I have called the unfolding of the leaf, the rate of elongation is faster and more or less linear until the leaf is nearly mature.

3. The time of unfolding of a leaf is very closely associated with the time of maturity of the next older leaf on the same side of the apex. Thus a leaf ceases growth when the next younger leaf immediately above it starts elongating rapidly, though which is cause and which is effect is it not possible to say.

4. This close relationship between duration of leaf elongation and rate of unfolding of successive leaves holds for both species in two seasons. It leads one to predict that selection for increased leaf size, in so far as it is a result of greater duration of leaf elongation, is likely to be accompanied by a slower rate of leaf appearance, and conversely that selection for rate of leaf appearance is likely to result in smaller leaves.

The kinetics of spread of colicin factor Ib (colIb) in col− cultures of S. typhimurium was studied. The rate of spread was greater with pil+ strains (those forming common pili) than with pil− strains. The difference reflects inefficient pairing between pil− cells, which donate or receive colIb efficiently only when mated with pil+ cells. The donor function of colIb is known to be repressed a few generations after it is acquired by a col− cell. While donor ability is manifest, a new type of pilus (the ‘Ib pilus’) is formed which is morphologically distinct from common pili and other sex pili such as that determined by the F factor. The Ib pilus is presumably involved in the transfer of colIb by conjugation.

Detailed measurements of leaf growth and leaf dimensions in the seedling stage were made on lines which had been selected either for large or small leaf size or for fast or slow rates of leaf appearance within one population of Lolium multiflorum (Italian ryegrass) and two populations of L. perenne (Irish perennial and Hunsballe perennial ryegrass).

Selection for either character had no effect on the rate of initiation of primordia at the apex, but did change the rate at which successive primordia became leaves. This rate of unfolding was very highly correlated with the rate of visible appearance of leaves and in all cases showed a parallel response to selection for the latter, as did also the rate of maturation of leaves. All three rates showed a negative correlated response to selection for leaf size.

Selection for increased leaf size in all cases led to a longer duration of the elongation of an individual leaf, but selection for faster rate of leaf appearance always reduced this duration. The rate of elongation of individual leaves increased under selection for larger leaf size but showed irregular changes under selection for faster leaf appearance, going down in Irish but up in Hunsballe.

Data for dimensions of cells from the lower epidermis showed that changes in leaf length under selection were sometimes associated with changes in cell length, some-times in cell number and sometimes with both.

Selection had in no case disrupted the close association between the maturation and cessation of growth of a leaf on the one hand, and, on the other, the unfolding from the apex and onset of rapid growth of the next younger leaf on the same side of the apex. Thus in all lines only two leaves (one on each side of the apex) were elongating rapidly at any one time, and an increase in the rate of unfolding was associated with a decrease in the duration of elongation and vice versa.

This association was the basis of the observed negative correlated responses between leaf size and rate of leaf appearance. But the fact that the rate of elongation could change independently of the duration opened up the possibility of setting up a selection criterion which would increase the total rate of leaf area formation.

The value of this kind of analysis of a character complex in a plant-breeding programme is suggested to lie in discovering physiologically or developmentally limiting processes rather than merely identifying morphological components.

Additional data are presented on the maintenance of kappa particles in cells of P. aurelia (stock 51) after removal of the gene K. It is shown that kappa particles may be maintained for as long as twenty-seven fissions in the absence of K, or may disappear before eight fissions. Slight retardation in fission rate of paramecia quickly eliminates kappa particles. The bearing of these results on the metagon theory is discussed.

1. A new semi-dominant gene called shaven (Sha) causes hairlessness in the mouse and it is closely linked to naked. Homozygous shaven mice never grow a first coat and adults grow a few short hairs only. Shaven heterozygotes grow a full coat which is greasy.

2. The percentage recombination between shaven and naked was found to be 0·8%. Three-point tests with naked and caracul showed that the order of the three loci is either N–Sha–Ca or N–Ca–Sha.

3. The failure of homozygous shaven mice to grow a coat is due to abnormal keratinization of the hairs in the follicles. The follicles were found to be deficient in sulphydryl groups.

4. The greasiness of the Sha+ coat is due to the presence of a sudanophilic fluid in the hair medullae.

5. The close linkage between the two genes with similar phenotypic effects is discussed.

Six lines have been isolated in which individual chromosomes of Ae. umbellulata have been added to the normal complement of T. aestivum. The phenotype and cytology of these lines and of the material employed in their production is briefly described. The use of some of these lines in investigations on the evolution of the diploid Aegilops species and in the introduction of useful alien variation into T. aestivum is mentioned.

It is generally accepted that reproductive isolation leading to the irreversible division of a Mendelian population into two species must be initiated during a period of geographical isolation (see Mayr, 1963). Thoday & Gibson (1962, Gibson & Thoday, 1963) obtained partial isolation within a population of Drosophila melanogaster by artificial disruptive selection for the number of sternopleural chaetae. This occurred in two experiments from the same base population after seven and twelve generations respectively. On the strength of these results they concluded that speciation does not require geographical isolation and that sympatric speciation by disruptive selection in a heterogeneous habitat is at least a theoretical possibility.

A lysogenic, recombination-deficient bacterial mutant (rec−) can be induced to produce λ ind− C587 by a temperature shift, although it does not produce phage after u.v. irradiation when lysogenic for wild λ. When this mutant (rec−) was made colicinogenic for colicins E2 and I, colicin production was found not to be inducible.

Superinfection of irradiated lysogenic rec− (λ) bacteria, at doses of irradiation that did not reduce their capacity to produce phage λ, was not productive. It seems therefore that a certain product(s) that usually arises following u.v. irradiation in rec+ cells, and triggers the induction of lysogenic and colicinogenic cells by causing loss of immunity or repression, is not produced in the rec− mutant. However, immunity is lost and the prophage is induced after a temperature shift in rec− cells which are lysogenic for λ ind− C857.

In their recent paper, Waddington & Robertson (1966) discuss the general importance of disruptive selection and make various points that call for comment.

First, their statement that ‘It is most desirable that Thoday's results … should be repeated, with special attention to the strict virginity of the flies used, …’ appears to some readers as a scarcely veiled imputation that my experiments, especially that of Thoday & Gibson (1962), were executed without technical competence, though Waddington has informed me such imputation was not intended.

A new mutant, Small eyes (symbol Sey), that reduces the size of the eye in the mouse, is described. It is dominant, and lethal when homozygous.

The influence on allelic recombination of non-lethal doses of gamma radiation delivered at different stages during meiosis has been determined and the data compared with previous results concerning gene recombination. Both kinds of recombination are influenced at only two, probably the same two, meiotic stages. During the first sensitive stage, probably in preleptotene, the frequency of allelic recombination is increased, but that of gene recombination decreased. The frequency of both is increased during the second sensitive stage, which is probably in pachytene.