This paper is concerned with three related aspects of the behaviour of populations under artificial selection for increased scutellar bristle number: (i) the pattern of response on the probit scale; (ii) the homeostatic behaviour of the selection lines on relaxation of artificial selection; and (iii) correlated responses in generation interval, reproductive capacity and competitive ability. The study was designed so that linkage would be a comparatively unimportant factor in promoting correlated responses to selection, and the effects of genetic sampling from generation to generation were also reduced to a low level.

Progress from the base mean of 4·05 bristles in females to a level of almost 8 bristles has been shown to involve two distinct phases with realized heritabilities of 0·34 and 0·10 respectively, the zone of transition corresponding closely to the position of the 6/7 threshold on the underlying scale. In addition to an apparent average reduction of about 25% in the additive genetic standard deviation in phase II by comparison with phase I, the loss in response due to the opposition of natural selection has been shown to reach a maximum near the zone of separation of the two phases.

The pattern of behaviour of the populations under artificial and natural selection has suggested the presence in the base population of genes of large effect on both bristle number and reproductive fitness. There is also evidence of additional genetic variation in bristle number which is effectively neutral with respect to fitness. Continued selection for increased scutellar bristle number in large populations has been shown to reduce mean competitive ability by more than 80%.

1. The effects of long-continued selection for body weight in two lines of mice, one large and one small, are described.

2. The large line showed a sharp increase in weight after remaining at an apparent limit for twenty generations. A rare combinational event is suggested as the most likely explanation.

3. Reversed and relaxed selection from the large line at the limit failed to yield any response. This indicates that effectively, the additive genetic variance in this line had been exhausted.

4. In contrast, the small line at the limit regressed slightly towards the base population when selection was relaxed. Reversed selection yielded a ready response until a new limit was apparently reached. Loci affecting body weight in this line had therefore not been fixed by selection.

5. Natural selection, operating on viability between conception and the time when the selection was made, appears to explain best the lack of fixation in the small line.

6. Attention is drawn to the necessity of more experimental work to elucidate the genetic nature of the limits to artificial selection.

1. Improved methods for the culture and preservation of Physarum polycephalum amoebae have been developed.

2. Strains of amoebae have been isolated which are stably resistant to 4, 8 or 16 μg./ml. actidione.

3. Sensitivity vs. resistance to 4 μg./ml. actidione is probably controlled by a single pair of alleles at a locus unlinked to the mating-type locus.

4. Plasmodia homozygous or heterozygous for the allele determining resistance at this locus are no more resistant to actidione than plasmodia homozygous for the allele determining sensitivity.

Derivatives of E. coli K12 which had received a number of different drug resistance factors by conjugation from Salmonella strains were found also to have acquired insensitivity to colicins. The insensitivity associated with some R factors was similar in character to immunity conferred by colicinogeny, while that given by others more closely resembled the higher level of resistance due to absence of the colicin receptor. It was, however, not complete. Many of the R factors also restricted the growth of various phages.

This paper describes the aspects of fertility that had been affected by selection on litter size. For twelve generations previously the mice used as parents were chosen because they had been born in large or small litters. At the end of this time, litters in the fertile strain averaged 11·1 young born alive, while the less fertile strain averaged 5·5.

It was found that male fertility and inherent viability of the young had nothing to do with the response although neither was excluded by the method of selection. Several contributions, however, were made by the females, who were affected not only in ovulation rate, but also in their control of pre-implantational losses, foetal mortality and mortality of the newly born.

Females from the less fertile strain were particularly prone to pre-implantational loss of eggs. It remains to be shown whether these were due to fertilizational or implantational failure.

The incidence of earlier and later embroyonic losses in females of the same strain were uncorrelated—Utters that were depleted early were neither more nor less inclined to be depleted later.

The cataract-producing mutant gene previously described as ‘Shrivelled’ appears to be a recurrence of the mutant described by Paget as cataracta congenita subcapsularis, and should henceforth be referred to as CatFr.

Curly-whiskers (cw) is a recessive gene which was found in 1958 by Mr C. J. W. Smith of the Chester Beatty Research Institute, London. It arose in a subline of the CBA/Cbi inbred strain. The first mutant animals were one male and one female in a litter of five. The two mutants were mated together and a sib-mated subline was continued from them in which 500 mice were bred, all of which were curly-whiskered. This established the mutant to be fully penetrant. Curly-whiskers resembles the hair-waving genes in causing waving of the vibrissae, but it has no obvious waving effect on the hairs of the coat. The coat texture is, however, slightly abnormal and Mr Smith noted also that on the CBA background there was an appreciable darkening of the coat colour. Homozygotes (cw/cw) are easily classifiable soon after birth by the curled vibrissae. Heterozygotes (+/cw) often have slightly curled vibrissae, and the gene is therefore not fully recessive; but the distinction between +/cw and +/+ could not be relied on, and in the linkage tests cw was treated as a recessive gene.

Argininosuccinase has been purified from wild-type Neurospora crassa, strain ST.A. The purified enzyme, which is homogeneous by the criteria of analytical centrifugation and starch-gel electrophoresis, has a molecular weight of about 175,000. The enzyme has also been partially purified from a heterokaryon between the arg-10 mutant stocks B317–9–9a and 402–3a.

The reaction kinetics of the two enzymes were compared in several respects, and they were found to be indistinguishable. The enzymes were also indistinguishable by starch-gel electrophoresis, and sedimented at the same rate through a sucrose gradient. It seems likely, however, that the enzymes do differ physically since they showed different affinities for both calcium phosphate gel and hydroxylapatite during purification.

The conclusion by Suit, Matney, Doudney & Billen (1964) that Hfr donor cells of Escherichia coli K12, starved of required amino acids can mate, has been re-examined. It appears that their conclusion is not valid and that apparent fertility of amino-acid starved cells is due to cross-feeding by the F− cells. The relationship of this result to the alternative mechanisms for chromosome transfer in E. coli is discussed.

In Cattanach's X-autosome translocation a piece of autosome of linkage group I has been inserted into the X-chromosome and a piece of X may have been reciprocally translocated to the autosome (Cattanach, 1961; Ohno & Cattanach, 1962). The present communication reports investigations to locate the autosomal insertion in the X-chromosome linkage map and provide evidence pertinent to the question of the possible reciprocal nature of the rearrangement; a brief summary of the results has already been reported (Cattanach & Isaacson, 1965).

The inducibility of three enzymes (β-galactosidase, tryptophanase and D-serine deaminase) has been measured at various times during the cell cycles of three strains of Escherichia coli (K12 58–161 F−, B/r F– and 15T−). In each strain sharp increases in inducibility of these enzymes occurred at characteristic periods in each cell cycle. Such increases depend on DNA replication and therefore probably reflect synchronized gene replication. It is inferred that chromosome replication in these F− strains is sequential from a fixed origin.

Infection with F′Lack+ results in an extra period of increase in inducibiity of β-galactosidase in each cell cycle. It is concluded that the F′ episome replicates once in each cell cycle at a time soon after cell separation.

A formula for effective population size when fertility is inherited is worked out. It is shown that the effective size decreases as the heritability of fertility or progeny number increases.