1. In a field experiment all plants of S. 48 timothy produced ears after being sown on or before 24 April; more than 90% of the plants still flowered if sown up to 19 May but fertility declined after later sowings, until only vegetative growth resulted in plants sown on or after 30 June.

2. In a pot experiment, during which the performance of individual tillers was recorded, all tillers arising after 23 July were found to remain vegetative. Ear-bearing capacity declined towards this date but was still high in tillers appearing as late as 2 July in plants sown in the middle of May.

3. Tillers on the main stem were more fertile and produced heavier ears than other tillers of the same age and order of succession on the plant.

4. Many plants sown late in the field produced ears which formed vegetative proliferations after the end of the summer.

5. The results are discussed in relation to the response of timothy to the changing environment, notably the length of day. It is suggested that timothy does not react to favourable photoperiods immediately upon germination and that failure to flower in plants sown after the middle of May does not necessarily imply any inadequacy of the environment to cause floral initiation, since it was possible for tillers appearing up to 2 July to produce ears, with few exceptions. Factors inherent in the organization of the plant are considered to be responsible for a decline of ear-boaring capacity from tiller to tiller, which occurred irrespective of changing environmental conditions.

1. Four varieties of marrow stem kale, three of thousand headed kale, Hungry Gap Kale, Rape-Kale and five other rape varieties were sampled at two centres in mid-Wales and separate leaf and stem samples analysed for manganese, molybdenum and boron. Similar samples of four varieties of marrow stem kale, two of thousand headed kale, Hungry Gap Kale, Rape-Kale and two other varieties of rape were taken from two further centres and prepared with the most stringent precautions to prevent contamination. Iron, copper, zinc and lead were determined in these samples.

An experiment to investigate the differential response of lambs of various breeds and crosses to different planes of nutrition is described.

Two groups of twin lambs were used comprising Blackfaces and their crosses ( x Lincoln, Wiltshire and Border Leicester rams) and Welsh and their osses ( x Suffolk and Wiltshire rams).

From October to January, each cross was divided between high and low planes of nutrition. Approximately half of the twin pairs were split and the remaining pairs distributed between the environments. Significant differences were found between the crosses and twin pairs of the same cross for body weight, body measurements and wool production but not for various blood characters. For almost all characters studied, plane differences were found. There was, however, only one significant (P < 1%) cross-plane interaction—for non-protein nitrogen concentration in the blood.

At the end of January, each plane was subdivided to produce high-high, high-low, low-high and low-low groups. The lambs remained in these groups until May. As before, differences between crosses were found for body weights and measurements, and wool characters but not for most blood characters. The majority of characters were affected by plane of nutrition. Cross-plane interactions were not found except in antibody response to Erisipelothrix rhusiopathia vaccine (P < 5%).

At the end of May the high-high group were slaughtered and the remaining three groups fattened on grass until each reached the same average weight as the high-high group. Carcass measurements showed that all the crosses shared a striking ability to recover from previous poor nutritional treatments. One significant (P < 5%) interaction of cross and plane was found in weight of cannon bone.

The most characteristic feature of the results has been the similarity of response of the different genotypes to the different nutritional environments. The few interactions found to be statistically significant by conventional methods are difficult to interpret because of the many tests of significance carried out in the analyses.

1. Each of three litter sister sows was fed one of three levels of crude protein in the diet during two 8-day periods in the fourth and seventh weeks of lactation. The three protein levels were 11, 15 and 19% of the diet on an air-dry basis. The scale of feeding was varied to maintain each sow at constant weight throughout the 8-week lactation period.

2. During two 3-day collection periods following 5-day preliminary periods, intake of nitrogen from the diet and outputs of nitrogen in faeces, urine and milk were measured.

3. The sow receiving the medium level of protein in the diet was in positive nitrogen balance during both collection periods.

4. The major effect of increasing the protein level of the diet was an increase in the output of urinary nitrogen. Neither milk yield nor output of protein in the milk appeared to be affected over the limited period studied.

5. Apparent digestibility of dietary protein was 75% on the low level and 85% on the medium and high levels of intake.

1. A theoretical feeding scale, designed to supply the full nutritional requirements of the lactating sow and prevent weight loss (allowing approximately 8 lb. meal per sow and 0·8 lb. per piglet), was tested against a much lower but commonly recommended level (2 lb. meal per sow and 1 lb. per piglet). The meal (barley meal, 40%; ground oats 25%; middlings, 20%; white fish meal, 7·5%;, groundnut meal, 2·5%; dried grass meal, 3·8%; ground limestone, 0·3%; common salt, 0·3%; vitamins A, D and B2 supplement, 0·6%) contained approximately 16·25% crude protein and had a digestible energy value of approximately 1300 Cal./lb. During pregnancies all sows were fed equal amounts of a meal containing approximately 13·0% crude protein.

2. Four pairs of Wessex Saddleback litter sisters were divided between the two treatments and records taken over three lactations of food consumptions, sow body weights, milk yields, milk compositions, litter growth rates, creep-feed consumptions and reproductive performances. Litter sizes were standardized between pairs of sows within the first 10 days after farrowing.

1. An in vitro apparatus is described for measuring the percentage dry-matter digestibility of roughages.

2. Samples of hay, dried herbage, oat hay and oat straw were tested in vitro and the values obtained compared with in vivo results obtained with sheep.

An account is given of the considerations which have been found in practice to govern the design of a small cabinet for growing plants under closely controlled conditions of atmospheric environment. Starting with questions of size, shape and material, the connexion between size of chamber and type of illumination is considered, followed by general policy on air supply. The main outlines of design having been laid down in this way, the various environmental factors to be controlled are reviewed. These include: (a) composition of the air, with particular reference to carbon dioxide; (b) light intensity. Various possible sources are reviewed, and the difficulties of imitating natural conditions of illumination are discussed, together with methods of measuring the illumination and checking for stability; (c) temperature. The degree of control needed for various purposes is considered, particularly in connexion with control of humidity, followed by systems of control, and the most advantageous arrangements for them. The cycle of operations of a control system is considered in some detail, and division of the system into a small relay-operated heater and a background heating or cooling system is advocated; (d) humidity. A similar division between background and relay-operated humidifiers is also advisable, and methods of achieving this are outlined. Dew-point control is shown to be most suitable for the background humidity, while a hot wick of low thermal capacity suffices for the relay-operated device. Finally, the principal uses of such cabinets are dealt with: (a) as adjuncts to field experimentation; (b) for work on plant pathology; (c) for producing standard plant material at any time of year; and rough estimates of running costs are given.

1. A crossing experiment was carried out between two breeds of sheep: the Wiltshire Horn—a breed with a short-stapled fleece which is normally shed both in lambs and adults, and the Scottish Blackface—which has a long-stapled carpet type of fleece. Animals of F1, F2 and both backcross generations were studied for gross characteristics of the fleece.

2. The influence of the Wiltshire genotype caused a high incidence of fleece-shedding animals amongst all the crosses. Three separate components of the shedding process were distinguished: (a) the ability to shed, (b) the extent of shedding—or proportion of the body area denuded, (c) the time of onset of shedding. Attributes (a) and (b) were determined genetically; control was probably on a multifactorial basis, although there may have been fewer genes involved than for the other characters studied. Given the genetically determined existence of (a) and (b), (c) was apparently controlled by the environment. Seasonal change in day length was postulated as a major operative factor.

3. The mid-side staple length of Wiltshire crosses was less than that of the Blackface parents. The decrease resulted partly from shedding and partly also from a reduction in the average growth rate of some fibres. When corrected for shedding the mean staple lengths of the crosses were about equal to the values theoretically expected on the basis of multifactorial inheritance without directional dominance.

4. The mean fleece weight of the Wiltshire crosses was less than that of the Blackface parents, due, about equally, to: (a) loss of wool by fleece shedding, (b) reduced wool production—probably caused, at least partly, by the decreased average growth rate of some of the component fibres of the fleece. Differences in wool production appeared to be inherited multifaetorially.

1. The availability of iron in five species of legumes has been investigated, using the rat as the experimental animal.

2. Two procedures were employed, namely, the increase in blood haemoglobin concentration, and the gain in total grams of haemoglobin, by nutritionally anaemic rats.

3. Inorganic iron, as ferric chloride, was shown to be significantly more available than the iron contained in the five legumes.

4. A significant difference in iron availability between alsike and wild white clover was obtained with both methods of expressing the results. Differences between other species were not significant.

5. The data are compared with those obtained for three species of grass in an identical experiment reported previously.

It was shown in previous work (Evans, 1958) that a cereal diet supplemented with 7% of white-fish meal supplied around 0·76% of lysine and 0·5% of methionine plus cystine. The addition of more pure l-lysine monochloride or dl -methionine to this animal-protein diet failed to bring about any significant improvement in the rate of growth, efficiency of food conversion or in the nitrogen retention of the pig, in the critical period between weaning and 80 lb. live weight. When the same cereal diet was supplemented by as much as 20% of ex. dec. ground-nut meal, however, the lysine content amounted to only 0·62% but the methionine plus cystine was 0·53%. On adding 0·2% l-lysine plus 0·2% dl-methionine to the 20% ground-nut diet the utilization of the nitrogen in the food was improved and the excretion of nitrogen in the urine was reduced. This finding was confirmed by means of a statistically designed growth trial, the daily rate of live-weight gain, improving from a mean value of 0·86 to 0·98 lb. following the addition of 0·2% of lysine and methionine, respectively, to the diet, the corresponding saving in meal consumption, between 35 and 80 lb. live weight, amounting to 16 lb. One group of ten pigs received a supplement of 0·2% of methionine only, and since it failed to show any improvement in performance over the control group it was concluded that 0·5% of methionine plus cystine must suffice to support satisfactory growth.

1. Field studies on Impatient parviflora DC. in shaded situations indicated the need for experimental investigation of the effects on growth of varying single factors of the aerial environment in otherwise constant conditions.

2. The use of cabinets which allow accurate control of the aerial environment for studies in physiology, morphogenesis, pathology and plant selection is discussed.

3. By comparison of the development and assimilatory ability of the leaves of I. parviflora, it is shown that plants may be grown in wholly artificial conditions which are substantially identical with those grown under comparable conditions out of doors.

The experiment with two trials was carried out to determine the effect of low and high germination and three seed rates of lucerne on the establishment, mortality, and yield of fodder on heavy soil at Cambridge during 1955–8.

Highly significant differences were obtained between seed rates and germinations in the number of plants established in the year of sowing. The highest number of plants with thin and short tap roots were in a dense crop at 15 lb./acre and 83% of them were unable to survive the first winter. The highest mortality occurred between sowing and the first harvest year, and the heaviest loss of plants was in the lucerne drilled at 15 lb./acre and lowest at 5 lb./acre. Interaction between germination and seed rates was significant at 1% level in the first and at 5% level in the second count only.

The highest differences in yield results were in the first harvest year, when a significantly lower yield was obtained at 5 lb./acre than at 10 or 15 lb./acre, but there was no significant difference in yield between 10 and 15 lb./acre. Lucerne with a high germination at all seed rates gave significantly higher yields than lucerne with a low germination in both trials.

It has been established that the mineral deficiencies can be rectified on both the fescue dominant and Molinia dominant swards of mid-Wales without recourse to ploughing and reseeding (Milton & Davies, 1947). The processes adopted consisted either of liming the native swards in the absence of fertilizers or else fertilizing the sward either in the presence or absence of lime. The grazing animal played an important part in the changes that took place. Some of the treatments resulted in the partial or complete extinction of the native hill herbage and its replacement by grasses and clovers of high productivity and mineral content.

An investigation of the estimation of the fertility of bulls' semen was carried out using the Impedance Bridge. A significant relationship between fertility and Impedance Change Frequency was confirmed, and this relationship extended to a within-bull significance.