A method of analysis of two factor experiments is given. This involves a development of the regression analysis used by Finlay & Wilkinson (1963) and others, and allows for the regression of interaction components onto both main effects. The usefulness of the single joint regression parameter for prediction is outlined.

The applicability of the analysis to three situations commonly found in grass breeding is illustrated by means of examples. It is concluded that the model may frequently describe variation due to genotype-environment interactions, and among diallel arrangements of binary mixtures of genotypes, but is likely to be of little utility for genetic diallel or other mating schemes unless the genes have a strongly correlated distribution among the parent plants.

Attempts were made to relate various parameters of growth in wheat substitution lines to final yield of grain and assess the extent to which such relationships could be useful as early predictors of yield. The associations were measured by calculating correlation coefficients, partial regression coefficients, standard partial regression coefficients and independent contribution to yield. Significant correlation coefficients were common but significant partial regression coefficients were rare. The remaining estimates provided a ranked order of characters associated with yield. Those parameters of vegetative growth showing most association with final yield were themselves related fairly directly to ear and grain size. It was concluded that direct estimates of yield must remain the most appropriate character on which to base selection in cereals.

The effects of defoliation regime on seasonal and total yields of spring- and autumnsown material of three populations Festuca arundinacea(syn. 1 and syn. 2 from North Africa and the British cultivar S. 170) were examined at Cambridge from the period June 1963 to April 1965.

A micro-plot technique was used in which defoliation treatments were defined by vertical height above soil level. The vertical growth intervals between successive defoliations were, for the summer cutting regimes 7.5–25 cm and 2.5–12.5 cm, and for the winter regimes 7.5–20 cm, 2.5–20 cm, 7.5–12.5 cm, and 2.5–7.5 cm.

In summer 1963 mean yields of all populations were very similar, being equivalent to approximately 12000 kg/ha. Mean population yields for the following three 6-month periods were equivalent to (in kg/ha): syn. 11519, 14740, 3385; syn. 2 2194,17669, 5523; S. 170 1656, 19832,3696.

Variety x treatment interactions were found in every season. In particular, S. 170 winter production was reduced by severe defoliations in the preceding summer, and syn. 1 and syn. 2 summer production by severe defoliations in the preceding winter. It is suggested that such interactions may have been responsible for conflicting assessments of the relative potential production of these populations.

Syn. 1 appeared to offer no advantage over S. 170, and syn. 2 only to outyield S. 170 where spring sowing was followed by infrequent or lax defoliations in all seasons. Agronomic implications are discussed.

1. A field trial in which a sward, predominantly of S. 24 ryegrass was fertilized with 248, 301, 532 or 589 kg N/ha showed that a maximal yield of 13·1 tonnes/ha of artificially dried grass could be obtained from four harvests in the year.

2. The materials harvested ranged in crude protein content from 10·1 to 23·6% and in apparent digestibility of dry matter from 72 to 76%. The amounts of the grasses voluntarily consumed by sheep were slightly and significantly lower for those containing the least N × 6·25.

3. In feeding trials with twenty-five steers, a barley diet was compared with a diet entirely of dried grass from the 3rd and 4th harvests from an area in the same field which had received 487 kg N/ha. The cattle given the barley diet consumed 8·7 kg/day and gained 1·22 kg/day. Those given dried grass consumed 8·9 kg and gained only 0·85 kg/day. Carcass quality of animals given dried grass was significantly lower than of those given the barley diet.

4. Calorimetric trials were made using two herbages from the first harvest (501, 532 kg N/ha) and all four herbages from the third harvest. In each of the six experiments three sheep were each given three amounts of a herbage, and in addition were fasted. Fifty-four calorimetric observations each lasting 5 days and eighteen fasting metabolism measurements were made.

5. With both 1st and 3rd harvests, methane energy losses fell and urinary energy losses increased with the level of N fertilization. Faecal energy losses were little affected by fertilizer application and the metabolizable energy (kcal/g D.M.) was 2·8 (11·7 kJ/g) for 1st harvests and 2·6 (10·9 kJ/g) for 3rd harvests. There was a slight decline in metabolizable energy/g D.M. with N application for the 3rd harvests.

6. The net energy values of the grasses for maintenance were 1·99 kcal/g D.M. (8·4 kJ/g) for the 1st harvested material and 1·84 kcal/g D.M. (7·7kJ/g) for the 3rd harvests. Net energy values for fattening, however, were 1·28 kcal/g D.M. (5·4kJ/g) for 1st harvests and 0·87 kcal/g D.M. (3·6kJ/g) for 3rd harvests. There were no significant effects of level of N fertilization.

7. The poorer nutritive value for fattening of the 3rd harvested grass, corresponding to an efficiency of utilization of metabolizable energy of only 33%, was contrary to prediction from equations relating nutritive value to either chemical composition or to metabolizable energy. The results of the calorimetric work, however, agreed with those of the feeding trial with cattle.

8. Combination of the results of the various trials showed that it was possible to obtain 1400 kg live-weight gain/ha from a system of husbandry based on high fertilizer use and artificial drying of herbage. With a barley system a maximal output/ha would be less than half this amount.

Casein in amounts ranging from 10 to 124 g/day was administered as an abomasal infusion to seven Merino wethers whose daily ration comprised 600–700 g of either chopped wheaten hay alone or a mixture of chopped wheaten and lucerne hays. Increases in both the length growth rate and diameter of the wool occurred at all levels of supplementation tested. When casein was infused at the rate of 100 g/day for a 20-day period the biggest increases in wool growth occurred during the first 4 days and there was little further change after the first 8 days.

There were marked wool growth responses to the infusion of up to 80 g casein per day. The results suggest that little further increase could be expected with the infusion of more than 120 g/day.

The performance of a series of intervarietal F1 hybrid populations was determined in simulated swards in a range of environmental conditions. On average the hybrids were more stable in performance than their parental varieties, but they showed no general yield advantage. Several hybrids were markedly heterotic at the higher-parental level, however. More heterosis was exhibited in high productivity conditions than in less favourable environments. A generally poor correlation between parental and F1 hybrid performance under competitive simulated-sward conditions and performance in noncompetitive spaced-plant conditions was found.

During the nineteenth century several parts of Wales were intensively mined for lead, zinc and copper ores. Fields adjacent to and downstream from the mines became contaminated by air- and water-borne heavy metal compounds. Such fields still contain high concentrations of total lead, zinc and copper together with silver and cadmium, the chief ‘guest’ elements in lead and zinc ores. Extraction of the soils with dilute acetic acid suggested that contaminated soils contained more of these metals that were available to plants (Alloway, 1969; Alloway & Davies, 1971). Some studies of plant composition were made to confirm the evidence from soil extraction and to indicate the extent to which these metals were entering the food chain.

Despite the fact that early embryo mortality commonly leads to the loss of 25% of fertilized ova in ewes (Edey, 1969), little is known of factors associated with this loss. Severe, acute undernutrition (Edey, 1966) and high ambient temperatures (Dutt, Ellington & Carlton, 1959) are specific factors which are known to cause embryo death, but these do not normally occur in most sheep producing areas.

The mechanisms responsible for moving K from soil to root were studied by growing perennial ryegrass in pots containing a flinty silt loam from Rothamsted and a sandy Woburn soil, with and without added K, at 0.75 water-holding capacity. More water was taken up from the Rothamsted soil than from the Woburn soil, and less was taken up when K was added without affecting dry-matter production. After 25, 52, 83, 122, 194 and 276 days, K uptake, water uptake and the K concentration in the soil solution were measured and the mass-flow contributionto K uptake calculated.

Without added K, mass-flow accounted for only about one-sixth of the K taken up from the Rothamsted soil and about one-third from the Woburn soil. Except at first, diffusion probably accounted for the remainder of uptake but was the rate-controlling step for uptake from the Woburn soil only. With added K, mass-flow became much more important relative to diffusion in both soils, and transported more than enough K to account for the measured K uptake. Root interception probably did not contribute to K uptake after the early part of the experiment, because the pots used were small.

Intravenous glucose tolerance tests were performed on thirty-four sows when they were either non-pregnant or in very early pregnancy. It was found that there was a relation between each sow's glucose tolerance and the mean of the mean birth weight of all her litters. The relation was expressed by the equation Y = 0.013T + 0.81 (R.S.D. ± 0–124 kg) where Y = birth weight and T = glucose tolerance as minutes for blood glucose to return to fasting concentration. The correlation coefficient was +0.58. The addition of further variables to the equation (sow's live weight at test, sow's parity at test and the number of pigs per litter) did not give any useful improvement of the estimate of birth weight.

A number of 237 Barki ewes of 2, 3, 4 and 6 years of age and weighing 25–45 kg were put under close observation so as to study the subjective and objective aspects of their behaviour during parturition.

During the choice and preparation of the parturition site ingestive behaviour ceased and an investigatory behaviour took place using the visual and olfactory senses. Ewes showed preference for the less illuminated and less disturbed places. They pawed the chosen spots with their hooves so as to prepare them for parturition. The parturient behaviour was characterized by signs of uneasiness and of pain, which increased gradually until the lamb's expulsion.

The overall average pre-partum time of labour was 46·1 plusmn; 1·6 min. The 2-year-old ewes took longer to give birth than older ones. The pre-partum labour was not significantly affected by body weights of ewes. The overall average duration of the post-partum labour was 143·27 min. Age and weight of ewes had no significant effect on the duration of the post-partum labour. Births were not evenly distributed over the period of observation.

The wool growth cycle in six rams and eight ewes was studied by taking monthly coat and skin samples for 3 years from each. Coat length was measured, and the percentage of inactive follicles was noted as well as the percentage of medullated fibres.

In the spring moult ‘mainland’ sheep (stock removed from St Kilda 60 years ago) cast less completely than ‘St Kilda sheep’ (Soays removed more recently from St Kilda). Mainland rams cast less than the ewes, but St Kilda rams began casting earlier, and completed casting more quickly than the ewes. Despite differences between years in the rate of casting, similarities in the pattern of casting were observed between years in the same animal, and between parent and offspring.

This apparent annual cycle was confirmed by graphs of follicle inactivity which showed that most inactivity occurred from September to March in the rams, and from October to April in the ewes. It was possible to fit sin and cos waves of 12-month duration to the graphs and further analysis revealed superimposed a 6-month cycle in one group, and a 4-month cycle in another group, which were associated with subsidiary peaks of inactivity in the summer. There were suggestions of a cycle about 2 months in length throughout the year, and sin waves of this duration could be fitted to the data from one of the groups of ewes. Possible control mechanisms for the annual and shorter cycles are discussed.

Peaks of follicle inactivity were frequently followed by troughs in coat length which suggested actual loss of the longest fibres.

Medullation was absent or minimal from December to March, and there were two peaks during the period of maximum medullation, that in the spring being greater than the one in the autumn.

Micro-organisms from the rumen of a hay-fed sheep rapidly synthesized an intracellular polysaccharide (starch) when growing or resting suspensions of cells were incubated in vitro with easily metabolized sugars.

In 30 min incubation periods the optimum pH for the synthesis of starch by resting cultures was about 6·0 when glucose or fructose were substrates. Relative to glucose (as 100%) in ability to form the polysaccharide were, fructose, 75%; sucrose, 80%; soluble starch, 18·6%; maltose, 6·9%; cellobiose, 4%; and xylose, 2·1%. No starch was formed from galacturonic, acetic, propionic, butyric, lactic or succinic acids. A bacterial fraction of the microbes was reponsible for about 80% of the starch formed from glucose, fructose or sucrose.

In incubations of 24 h, resting cultures formed more starch per unit of microbial protein than growing cultures. The utilization of microbial starch and lactic acid, formation of which often accompanied starch synthesis, gave rise to volatile fatty acids. Acid production was maintained from these substrates at rates similar to those obtained from the fermentation of glucose. The acids were in molar proportions of 65–70% acetic, 20–27% propionic and 8–15% butyric. The maximum starch calculated to be synthesized by the microbes from 100 ml of rumen liquor, in media containing excess sugar, amounted to over 250 mg from glucose, 200 mg from fructose, 200 mg from cellobiose and 50 mg from xylose. It is calculated that under optimum conditions for synthesis about 25 g of starch would pass daily from the rumen of a sheep.

Observations were made on the development of the testes and epididymides of thirty-two Clun Forest ram lambs. There were no significant differences between the weights and volumes of the left and right testes and epididymides.

The rate of increase in testicular and epididymal weights was linear up to a body weight of 46 lb (20·9 kg) followed subsequently by a significantly greater rate of increase at higher body weights. Similar patterns of growth occurred in relation to age, although at a given age there was a wide range of testicular and epididymal weights. There was a close relationship between the development of the testis and the epididymis at body weights greater than 46 lb (20·9 kg).

There were two distinct phases in the dimensional growth of the testis. Increases in testicular length and breadth proceeded at much the same rate at testis weights of below 20 g. During the subsequent phase testicular length increased at a significantly greater rate than testis breadth.

In an incomplete block 2x2x2 factorial experiment seventy-two Large White type pigs, after having been fed individually and identically up to 120 lb (54·5 kg) live weight, were fed from that weight up to 200 lb (90·9 kg) live weight maize or barley based diets (factor 1), with each diet having either a wide or a narrow calorie/protein/lysine (CPL) ratio (factor 2). The feeding was effected in individual feeders in a Danish type piggery (with six pigs per pen) with all diets offered as wet mashes and according to scales which gave two different calorie intakes (factor 3). There were, therefore, four different diets and eight different feeding treatments.

Calves were slaughtered at predetermined weights after being fed individually on fresh and reconstituted whole milk at three levels. The highest level of feeding was ad libitum and the lowest was aimed at keeping the calves at their birth weight for 72 days before providing extra milk to allow growth to proceed.

Dressing percentage did not decline with age as it did in a previous experiment using roughage in the diet. The faster-grown calves had a higher dressing percentage. It is concluded that both rate of growth and the physical nature of the diet influenced dressing percentage.

Muscle: bone ratio and the amount of fat and bone in the carcass were not significantly affected by the rate of live-weight increase.