1. Whole-milk powders containing 25 g water/kg were stored for up to 9 weeks in sealed aluminium containers at elevated temperatures. Lysine and other essential amino acids were measured by chemical and microbiological methods.

2. Storage at 60° resulted in the progressive formation of lactulosyl-lysine. After 9 weeks, 30% of the lysine groups were present in this form. The powders still retained their natural colour and the levels of tryptophan, methionine, cyst(e)ine and leucine remained unchanged.

3. Storage at 70° resulted in the formation of lactulosyl-lysine followed by its complete degradation with the development of browning. Available tryptophan, methione, leucine and isoleucine decreased progressively during storage.

4. The different methods for lysine determination gave widely dissimilar results. The direct fluorodinitrobenzene (FDNB) technique and reactive lysine from furosine were considered to be the most reliable methods. The FDNB-difference, dye-binding lysine, Tetrahymena and Pediococcus methods all seriously underestimated reactive or available lysine in heat-damaged milk powders. Tetrahymena and Pediococcus appeared to utilize lactulosyl-lysine as a source of lysine.

5. The results are discussed in relation to storage and distribution of milk powders in hot climates.

1. A questionnaire inquiring about food intolerance was sent to a random sample of the electorate in a small South Wales town. Replies were received from 170 men and 305 women, the response rates being 87 and 93% respectively.

2. Adverse effects of some food were reported by 19% of men and 26% of women, or 14 and 18% respectively if minor digestive symptoms are excluded. Certain foods were thought to cause non-abdominal symptoms by 4% of men and 10% of women, and vomiting, diarrhoea or abdominal pain by 11% of men and 10% of women.

3. Plasma IgE was measured in a random subset of ninety-nine women and found to be significantly lower in those with major symptoms than in the rest. This suggests that allergy is probably not a common cause of food intolerance.

1. Pigs weaned at 2 d of age were given diets in which a functional fish-protein concentrate (FFPC) supplied 0 (diet U), 350 (diet H), 525 (diet I) or 700 (diet G) g crude protein (nitrogen × 6·25, CP)/kg total CP. The remainder of the CP was supplied by dried skim-milk and whey. The diets were reconstituted (200 g dry matter (DM)/l), from two spray-dried powders (diets U and G) and these were mixed in the proportions 1:1 and 1:3 (v/v) to give diets H and I respectively. All four diets were given at hourly intervals on a scale based on live weight, and supplemented with vitamins and minerals. All diets contained the same amounts of calcium, phosphorus, sodium and potassium.

2. In an experiment to 28 d of age with ten pigs/diet, performance was inversely related to the amount of FFPC in the diet. One pig (diet I) and four pigs (diet G) died after scouring. In healthy pigs, N retention (g/d per kg live weight) was not affected by partial replacement of dried skim-milk (diet H).

3. Digestion of protein was studied in a further experiment with eight pigs/diet, killed at 7 d of age and at 1 h after a feed. The amount, pH, DM and total N contents of the digesta in the stomach were also inversely related to the amount of FFPC in the diet.

4. FFPC had no consistent effect on proteolytic enzymes. Greater amounts of trypsin and chymotrypsin were present in the distal compared with the proximal region of the small intestine, but source of dietary protein had no effect.

5. The proportion of non-protein-N in the total N in digesta from the small intestine was lower in pigs receiving FFPC suggesting that this protein may be more resistant to proteolysis than milk proteins. The apparent absence of an increase in enzyme secretion when FFPC was given could, therefore, explain the decrease in performance.

6. Some general effects of non-milk proteins from this series of experiments in the baby pig are discussed. It is suggested that performance may be inversely related to the rate of gastric emptying, as indicated by the amount of DM in the stomach. Unlike milk, other sources of protein did not coagulate in the stomach, and reduced the pH, DM and total N contents of the digesta. Giving diets containing a single-cell protein did appear to stimulate proteolytic enzyme secretion, unlike fish or soya-bean proteins.

1. Storage of milk powder under unfavourable conditions accelerates the normally slow deterioration in nutritional quality. The effects of such storage on the water-soluble vitamin composition were examined.

2. (a) Spray-dried whole milk containing 25 g water/kg was stored at 60° and 70° and sampled weekly to 9 weeks. (b) Spray-dried whole milk and skimmed milk were adjusted to contain 40 and 100 g water/kg and stored at 37° in nitrogenand in oxygen. Samples were taken for analysis at intervals during storage.

3. The samples were analysed for eight B-complex vitamins and ascorbic acid, and also for total lysine, ‘reactive lysine’ and ‘lysine as lactulosyl-lysine’.

4. Storage at 60° caused rapid destruction of folic acid (53% loss at 4 weeks) and slower loss of thiamin, vitamin B6 and pantothenic acid (18% at 8 weeks). There was no change in the content of riboflavin, biotin, nicotinic acid and vitamin B12. At 70° the rate of destruction of the four labile vitamins was much increased; 18% or less survived at 4 weeks.

5. At 37° and 40 g water/kg there was little change in total and ‘reactive’ lysine during storage for 57 d. Lactulosyl-lysine was demonstrably present butatlow concentration. There was considerable loss of folate (72%) and ascorbate (91%) during storage for 30 d in O2, but no significant loss in N2. Thiamin fell by approximately 12% in 57 d, equally in O2 and N2. The content of the remaining vitamins was unchanged. At 100 g water/kg there were progressive Maillard changes. During 27 d in N2 the colour changed from cream to palebrown, but in O2 there was no perceptible colour change. Total lysine fell by 20% in 27 d, and ‘reactive lysine’ by 30%. Folate was stable during 16 d in N2, but largely (94%) destroyed in O2. Ascorbic acid was also destroyed in N2 as in O2. Thiamin fell by 41% in 27 d, equally in O2 and N2. Vitamin B6 was more labile, especially in N2, falling by 71% in 16d.

6. With skimmed-milk powder containing 100 g water/kg, storage at 37° in O2 and N2 gave much the same results as for the corresponding whole-milk powder. The presence of milk fat had no marked effect on the stability of the water-soluble vitamins.

7. Destruction of vitamins was clearly linked to the progress of Maillard-type reactions and was strongly influenced by time and temperature of storage, moisture content and, in some instances, by the presence of O2.

1. Mature Merino ewes were given either a low-zinc diet (4 mg/kg) or an adequate-Zn diet (50 mg/kg) for all or part of pregnancy.

2. The ewes consuming the low-Zn diet consumed 25% less feed than those given the adequate-Zn diet during the last 115 d of pregnancy.

3.Zn concentration in the plasma of Zn-deficient pregnant ewes declined from 0·7 to 0·3 mg/1.

4. The lambs born to Zn-deficient ewes weighed less and had reduced concentrations ofZn or less total Zn, or both, in the whole carcass, liver and pancreas.

5. A reduction in activity of alkaline phosphatase (EC 3.1.3.1) in the liver and a slight reduction in thymidine kinase (EC 2.7.1.21) activity in the thymus was also observed in Zn-deficient lambs.

6. The Zn-deficient ewes deposited approximately 63 mg Zn into each single-born lamb;this indicates that during the last third of pregnancy the developing foetuses were accumulating the equivalent of 35% of the total dietary Zn intake of the ewes.

1. The effects of level of feeding of either a protein-adequate or a protein-deficient diet on nitrogen retention (NR), growth performance, body composition and some aspects of energy utilization in pigs growing from 1·8 to 6·5 kg live weight (LW) were investigated in two experiments.

2. In Expts 1 and 2 piglets were given a protein-adequate diet at four levels of intake (0·93, 1·44, 1·83 and 2·30 MJ gross energy (GE)/kg LW0·75 per d) and a protein-deficient diet at five levels of intake (1·14, 1·38, 1·68, 1·95 and 2·30 MJ GE/kg LW0·75 per d) respectively.

3. For pigs given the protein-adequate diet (Expt 1) NR was linearly (P < 0·001) related to energy intake (EI) and independent of N intake (NI). NR in pigs given the protein-deficient diet (Expt 2) was linearly (P < 0·001) related to NI and independent of EI.

4. Average daily LW gain responded linearly to increases in EI in both experiments. However, at equivalent levels of EI pigs given the protein-adequate diet exhibited more rapid and efficient growth than those given the protein-deficient diet. The results also indicated an interaction between the effects of EI and dietary protein content for feed conversion efficiency.

5. Body fat at 6·5 kg LW increased in a curvilinear fashion with increasing EI in both experiments. However, over the range of EI tested (from approximately 1·8 to 4·6 times energy for maintenance) body fat increased by 153% in Expt 1 and by only 27% in Expt 2. Pigs given the protein-deficient diet were also markedly fatter than those given the protein-adequate diet. Body protein at 6·5 kg LW decreased (P < 0·01) with increasing EI in Expt 1 but was unaffected by EI in Expt 2.

6. As estimated by multiple regression analysis, the values for the efficiency of energy utilization for protein (kp) and fat (kf) deposition were 0·76 and 0·78 respectively in Expt 1 and 0·42 and 0·89 resepctively in Expt 2. The estimates of metabolizable energy required for maintenance were 445 and 532 kJ/kg LW0·75 per d for Expts 1 and 2 respectively.

1. Two experiments are reported. In Expt 1, pregnant Sprague–Dawley rats (200–220 g) were given purified diets containing 35, 75, 150 or 300 mg Fe/kg throughout gestation and lactation. In Expt 2, the levels of Fe given were 250 and 300 mg/kg diet.

2. Litters were standardized at birth to contain seven pups. On day 20 of lactation blood, milk, spleen and liver were collected for Fe analysis.

3. A dietary Fe concentration of 35 mg/kg maintained maximum weight gain, food intake, liver and spleen weight, haemoglobin concentration and packed cell volume in the growing maternal rat through 20 d of lactation. Dam liver and spleen Fe concentrations were maximized between 75 and 250 mg Fe/kg diet.

4. Pups (20-d-old) nursed by dams given 35 mg Fe/kg tended to have lower body and organ weights compared to other groups. Liver and spleen Fe concentrations increased with increasing maternal dietary Fe and were significantly increased between 150 and 250 mg Fe/kg.

1. In three experiments, mature Blackface wethers were given freeze-stored Agrostis festuca herbage by continuous feeder.

2. In Expt 1, on separate occasions [U-14C]threonine, [U-14C]glucose and NaH14CO3 were infused over 12 h periods to obtain estimates of irreversible loss rate (ILR) of threonine, glucose and carbon dioxide in the plasma and of the exchange of C between these metabolites.

3. In Expts 2 and 3, during periods when glucose and threonine metabolism were examined, glucose loss across the kidneys (23–29 g/d) was induced by infusion of phloridzin.

4. Results from the four sheep used in Expts 1 and 3 are presented as three-pool models. They indicate that threonine ILR (7·8 g/d; 3·1 g C/d) was approximately three times the estimated rate of absorption of exogenous threonine (1 g C/d). Glucose ILR was approximately 76 g/d (mean ± SE; 30·3 ± 0·57 g C/d). Only 0·3% of the glucose-C (0·09 g/d) was derived directly from threonine-C (i.e. 3% of the threonine-C ILR). Bicarbonate ILR was 170 ± 7·3 g C/d, and glucose contributed 11·1 ± 3·52 g C/d to this, accounting for 51 ± 4·4% of glucose-C ILR. Threonine contributed 0·20 ± 0·026 g C/d to the bicarbonate-C ILR, accounting for only 6·4 ± 0·87% of the threonine-C ILR.

5. When, in Expts 2 and 3, phloridzin was infused, glucose ILR was increased by 28 ± 1·5% and bicarbonate ILR was increased by 13 ± 2·4%. Threonine ILR (3·1 g C/d) was not increased, but the metabolic distribution of threonine-C was altered. The transfer of threonine-C into glucose and CO2 was increased by 39 and 69% respectively to 0·125 and 0·45 g C/d, accounting for 4 and 13% of the threonine ILR respectively.

6. Both technical and metabolic considerations which affect interpretation of these results in terms of rates of catabolism of threonine and of quantitative estimates of gluconeogenesis from threonine are discussed.

1. Weanling rats fed on a relatively low protein diet were infected with several intestinal tapeworms. The weight gain and daily food intake of the rats were measured for 3 weeks before the animals were killed and the adult tapeworms recovered. The tissues of the rats and worms were then analysed for protein. Controls were provided by uninfected rats either pair fed or fed ad lib. For comparison, a similar experiment was conducted on rats fed on a relatively high protein diet.

2. The effect of the infection on the protein-malnourished rats and of the low level of protein on the worms were apparently not significant. The amount of protein contained in the worm burdens was less than 1·5% of the average total intake of the protein-malnourished rats.

1. The influence of trace-nutrient-binding proteins on the growth of coliforms, streptococci and lactobacilli in the gastrointestinal tract was examined in neonatal rabbits delivered germ-free and dosed with an artificial flora (ESL), or born conventionally and dosed with ESL or rabbit faeces.

2. In the stomach and small intestine of both gnotobiotic and conventional animals the counts of coliforms were usually atypically high and those of streptococci were always low. In the colon the counts of coliforms and streptococci were high. Lactobacilli usually became established in the gut of the gnotobiotic animals but were not found in the conventional rabbits.

3. Sterilization (freeze-drying followed by γ-irradiation) of the milk decreased its capacity to bind added iron by 45% and vitamin B12 by 30%. When compared with raw milk, feeding of radiation-sterilized milk did not affect the viable count of coliforms and streptococci in the gut of gnotobiotic animals.

4. Saturating the nutrient-binding proteins in milk with Fe, folic acid and vitamin B12 had no effect on the numbers of coliforms, streptococci and lactobacilli recovered from the intestine.

1. The irreversible loss rate (ILR) of glucose, bicarbonate and threonine were determined in six twin- and triplet-bearing ewes on three occasions during the last 6 weeks of gestation.

2. Three of the ewes (group S) were given conventional rations of hay plus concentrates so that their blood ketone levels did not rise over this period. The other three ewes (group H) were given a fixed intake of hay throughout; their blood ketone levels rose, but remained at subclinical levels.

3. The results are presented in the form of three-pool open-compartment models for each period. There was an increase in the glucose ILR for both groups over the 6-week period, but the over-all increase (mean ± SE) was much greater (p < 0·001) in group S (35·1 ± 2·43 g carbon/d) than in group H (11·3 ± 1·28 g C/d). Similarly, increases in bicarbonate ILR were also higher in group S (161 ± 11·2 g C/d) than in group H (63 ± 21·7 g C/d). However, whereas with group S ewes, receiving supplementary feed, this increase was progressive throughout the 6 weeks, with the group H ewesthe increase which occurred over the last 3 weeks of gestation (56 ± 26·3 gC/d) was much greater than that which occurred over the preceding 3 weeks (7 ±4·7 g C/d). This pattern was also evident in the oxidation of glucose to CO2.

4. In contrast to the previously mentioned findings, neither threonine ILR nor the amount of threonine converted to glucose or catabolized to CO2 changed significantly over the 6-week period.

5. The results are discussed in the light of findings presented in the previous paper that the amount of threonine used in catabolic processes can alter if a glucose-only sinkis created in wether sheep.

1. In an attempt to investigate calcium absorption in the rat during the postprandial period, with the least alteration of the physical environment, the undisturbed small intestine was ligated in situ 2·5 or 3·0 h after ingestion of a diet containing 200 g casein/kg or an equivalent amino acid mixture, or 925 g casein/kg. Estimation of Ca absorption was made by comparing the amount of soluble 40Ca or 45Ca in the contents of segments from the rats receiving 45Ca by intubation 30 min after withdrawal of food, ligated after a further 30 min, then killed after 0 or 30 min.

2. Under conditions such that the estimated amount of a marker, polyethylene glycol, in segments ligated in a defined position was little changed in rats killed 30 min apart, the difference in the amount of soluble 40Ca was much higher in the rats fed on the basal diet containing 200 g casein/kg than in other groups.

3. This specific effect on Ca absorption, particularly in the distal portion of the small intestine, could be seen also after 45Ca was directly injected into ligated segments in situ. The amount of 45Ca in the portal blood 15 min after injection of the label was also highest in the rats given the basal diet.

4. The results were in agreement with our previous findings that the formation and accumulation of casein phosphopeptides causes an increase in the amount of soluble Ca in the distal small intestine.

1. Kale (Brassica oleracea) and ryegrass (Lolium perenne)–clover (Trifolium repens) pasture, grown under similar soil conditions, were grazed in the vegetative state by growing lambs of 23·6 kg initial live weight for 24 weeks. The kale and pasture contained respectively 20 and 270 μg iodine/kg dry matter (DM). The kale also contained 8 μmol total glucosinolates/g DM and 11·5 g S-methyl-L-cysteine sulphoxide (SMCO)/kg DM, both of which were nondetectable in the pasture diet.

2. Intramuscular injections of 1 (475 mg) were given during weeks 1 and 12 to half the forty-eight lambs grazing each forage. Wool growth, live-weight gain and cytochrome oxidase (EC 1.9.3.1) activity of biopsied hind-limb muscle were measured at 6-week intervals. Jugular blood samples were removed every 6 weeks for the determination of haematological factors and serum thyroid hormone concentrations. All animals were slaughtered at the end of the experiment and thyroid weight, thyroid I content, and the weight and cytochrome oxidase activity of heart muscle determined.

3. Serum concentrations of thyroxine (T4) increased from 20 to 48 nmol/l during the 24 weeks that control lambs grazed ryegrass-clover pasture. I supplementation increased the concentration and total amount of I in the thyroid gland and increased serum T4 concentration, but did not affect any other values measured in the lambs grazing the pasture herbage. Serum concentrations of triiodothyronine (T3) were stable at 2 nmol/l for both groups.

4. Control lambs grazing kale for 24 weeks showed marked thyroid enlargement and depletion of thyroid I. By week 6, serum T4 and T3 concentrations had declined to 2–5 nmol/l and 1 nmol/l respectively and were stable at these values for the remainder of the experiment. I supplementation eliminated the thyroid depletion of this element, caused serum T4 concentration to rise and stabilize at 90 nmol/l by week 18, and T3 concentration to stabilize at 2 nmol/l by week 6. From week 6 onwards, wool growth was increased 13% by I supplementation, whereas empty body growth was unaffected.

5. Lambs grazing kale developed haemolytic anaemia, due to rumen fermentation of SMCO. I supplementation enabled the lambs to resist the anaemia better by increasing erythrocyte reduced gluthathione (GSH) content. Relative to pasture-fed animals, lambs grazing kale and supplemented with I showed increased heart muscle weight and cytochrome oxidase activity. This represented a compensatory mechanism for the reduced blood oxygen-carrying capacity caused by the anaemia. I-deficient (i.e. control) lambs grazing kale showed reduced cytochrome oxidase activity in both heart and hind-limb muscle.

6. The findings are in accord with T3 having a greater biological potency than T4 for regulating rates of body and wool growth. Increases in heart weight, heart cytochrome oxidase content and erythrocyte GSH content of kale-fed lambs were, however, associated with elevation in serum T4 and not T3 concentration.

7. I requirements of growing sheep and cattle consuming the pasture diets are discussed. Because of its better relationship to production traits, it is considered that requirements should be based on the ability to maintain T3 rather than T4 concentrations. On this basis, requirements could be met by diets containing 180–270 μg I/kg DM.

1. The heat losses and energy and nitrogen balances of thirty-six individually-housed, entire male pigs (initial body-weight 18–30 kg) were measured over 7 d periods, when they were fed on rations containing 153, 201 and 258 g crude protein (nitrogen × 6·25; CP)/kg dry matter (DM). The rations also contained 16·29, 16·96 and 17·24 MJ metabolizable energy (ME)/kg DM so that the CP:ME values were 9·4, 11·8 and 15·0 g CP/MJ ME respectively. Each ration was given at three levels, 20, 35 and 50 g feed/kg body-weight per d, thus giving nine dietary treatments. The experiments were carried out at an environmental temperature of 22 (±1)°.

2. Heat loss (H) increased significantly (P < 0·01) with increase in ME intake. The rate of increase in H was not, however, influenced by the protein content of the ration. Thus, energy retention (ER) at any given level of ME intake was independent of the ration offered. From the relationship between ER and ME, estimates of the maintenance energy requirement (MEm) and the partial efficiency of energy utilization (k) were determined. MEm varied within the range 494–568 kJ/kg body-weight0·75 per d, while k varied from 0·70 to 0·76.

3. Both energy and protein intakes had a significant influence upon the rates of protein (P) and fat (F) deposition, and hence body-weight gain. At any given level of feed intake P was higher and F lower the higher the protein content of the ration. However, when compared at similar levels of protein intake, both P and F were reduced the higher the protein content of the ration.

4. From the multiple regression equations relating P and F to ME, individual estimates of MEm and the energetic efficiencies of protein (kp) and fat (kF) depositions were determined. Using an overall mean kF value of 0·86, it was calculated that MEm ranged from 462 to 525 kJ/kg body-weight0·75 per d while kp varied from 0·48 to 0·55. The significance of these estimates of kp are discussed in the light of their derivations and in relation to theoretical values.

1. The apparent absorption values of individual amino acids from two samples of feather meal (FM) were determined in the lower ileum of chicks fed on diets containing magnesium ferrite as a marker.

2. The average absorption values for FM amino acids were low, approximately 0.50, as compared with approximately 0.85 for soya bean, used as a control. Values for individual amino acids from FM differed distinctly, ranging from 0.20 to 0.70. Low values were obtained for aspartic acid, histidine, lysine, glutamic acid and cystine.

3. An indication of the low absorption of the previously-mentioned amino acids was obtained by analysing the amino acid composition of the FM residues undigested by pepsin or pancreatin.

4. The reasons for testing the apparent rather than the true absorption are discussed.

1. Twenty heifer calves were allocated to four groups and maintained for 32 weeks on a diet based mainly on barley and straw and containing 4 mg copper/kg. The diet was supplemented with 0 or 800 mg iron/kg and 0 or 5 mg molybdenum/kg.

2. Liver and plasma Cu concentrations, erythrocyte superoxide dismutase (EC 1.15.1.1) and plasma caeruloplasmin (EC 1.16.3.1) activities decreased greatly and rapidly in all calves given the Fe or Mo supplements or both. Levels indicative of severe Cu deficiency were attained within 16 weeks. There were no significant differences in values in animals given Fe, Mo or Fe plus Mo.

3. Clinical signs of Cu deficiency developed after 20 weeks in the calves given the Mo supplement. Growth rates were reduced, skeletal lesions developed and hair texture and colour were affected. No such effects were observed in the calves given only the Fe supplement.

4. Plasma and liver Fe concentrations increased in calves given the Fe supplement but were not greatly affected by Mo, even when the calves were severely Cu-deficient.

5. The significance of the effects of Fe and Mo on Cu metabolism are discussed with special regard to the influence of soil ingestion on Cu availability and to the frequent lack of correlation between the Cu status of animals and their clinical condition.

1. The nitrogen balances of thirty-six individually-housed, entire male pigs (body-weight range 19–50 kg) were measured over 7 d periods when the animals were kept initially at an environmental temperature of 22° and then at 10° while fed on rations containing 153, 201 and 258 g crude protein (N × 6·25; CP)/kg dry matter (DM). The respective metabolizable energy (ME) contents were 16·29, 16·96 and 17·24 MJ/kg DM. Each ration was given at three levels, 20, 35 and 50 g feed/kg body-weight per d. The animals fed on the 20 and 35 g/kg feeding level were catheterized and blood samples withdrawn on two consecutive days within the N-balance periods for the determination of blood urea (BU) concentration both before and at hourly intervals for 7 h following the morning feed.

2. An increase in feed intake resulted in a significant increase in N retention (NR) at each environmental temperature. However, NR as a proportion of N intake was higher the lower the protein content of the ration. With the exception of the animals fed on the low-protein ration, NR at any given feed intake was lower at 10° than at 22° and these differences were reflected in the animal's body-weight gain.

3. Values for the fasting N metabolism (Nf), calculated from the relationship between NR and intake of digestible N (IDN), were temperature-dependent. At 22°, a constant Nf value of 0·255 g N/kg body-weight0·75 per d was found appropriate, while at 10° Nf increased with increase in protein content of the ration from 0·380 on the low protein ration to 0·533 and 0·753 g N/kg body-weight0·75 per d on the medium- and high-protein rations respectively.

4. The efficiency of N utilization (kN) reflected the differences in the relationships between NR and IDN. At 22° the relationship was curvilinear so that kN decreased with increase in both the level of feed intake and the protein content of the ration. At 10° the relationship was linear, hence kN was independent of feed intake within rations. However, it decreased from 0·909 to 0·679 as the protein content of the ration was increased.

5. The concentration of BU attained a maximal value some 3–5 h after the ingestion of the feed, with the values at 10° being higher than those at 22°. BU increased as the level of protein in the ration increased but decreased with the level of feed intake when dietary protein concentration was held constant. There was a significant correlation between BU and kN, indicating that BU is a useful criterion for assessing the efficiency of N utilization.

1. Six pigs, initially of 35 kg mean live weight, were each fitted with a re-entrant cannula. This was formed on either side of a short pouch of duodenum into which the pancreatic duct opened and which contained a simple cannula linked to the centre of the re-entrant cannula.

2. Each pig received two diets: diet A was based on wheat starch, sucrose and casein, while diet B was based on barley and soya-bean meal. The diets were given in equal amounts at 12 h intervals.

3. Digesta and pancreatic juice were collected continuously during three 12 h periods for each pig on each diet.

4. Mean duodenal output: dietary intake values for diets A and B respectively were: digesta 1·80, 2·86; dry matter 1·05, 1·03; nitrogen 1·05, 1·06; trichloroacetic acid (TCA)-soluble N 7·69, 9·10; glucose 0·97, 0·89. For diet A the proportion of TCA-soluble N in total N rose from 13 to 50% during 12 h, while it was approximately 50% throughout 12 h for diet B.

5. Mean total pepsin (EC 3.4.23.1) activities (units/24 h) were 760449 (diet A) and 1466571 (diet B).

6. Salivary and gastric secretions were calculated to be approximately 4 and 8 kg/24 h for diets A and B respectively.

7. Mean flows in pancreatic juice (g/24 h) for diets A and B respectively were: juice 1204, 2182; protein 10·94, 12·10; N 1·98, 2·14; ash 9·46, 17·31; sodium 3·88, 6·91; potassium 0·23, 0·54: calcium 0·031, 0·046; phosphorus 0·024, 0·026.

8. Mean total enzyme activities (units × 10−3/24 h) for diets A and B respectively were: trypsin (EC 3.4.21.4) 138, 114; chymotrypsin (EC 3.4.21.1) 84, 84; carboxypeptidase A (EC 3.4.2.1) 5, 4; carboxypeptidase B (EC 3.4.2.2) 15, 17; amylase (EC 3.2.1.1) 1061, 981.

9. It was calculated that the minimum amount of endogenous N from saliva and gastric secretion was 0·3–0·6 g in 24 h. This assumes no absorption of N occurred anterior to the duodenal cannula.

1. Studies were made of the extent to which p-cresol, catechol, quinol and orcinol infused through rumen or abomasal cannulas to sheep were recovered in their urine.

2. Rumen fermentation of dietary phenolic compounds caused the excretion of simple phenols in the urine. In decreasing order of magnitude these were: p-cresol, catechol, phenol and 4-methylcatechol with only traces of quinol and orcinol.

3. The percentages of rumen-infused p-cresol or orcinol recovered as increments in the urinary phenol outputs of sheep (94 and 99% respectively) following infusion showed that rumen degradation of these phenols was negligible.

4. After rumen infusion of catechol and quinol, mean recoveries of these phenols in urine were only 55 and 77% respectively. Possible reasons for these incomplete recoveries are discussed.

5. Studies were also made of the use of the urinary phenol output of phenols characteristic of particular forages as indices of their voluntary intake by sheep. Calluna vulgaris L. (Hull) (heather) may contain 1300–3600 mg/kg dry matter (DM) of orcinol and 200–800 mg/kg DM of quinol as β-glycosides. When heather was offered ad lib. to sheep given one of five levels of grass, linear relationships were found between heather intake and urinary quinol and orcinol outputs.

6. The urinary output of aromatic acids was also determined when sheep ate grass and heather. Urinary phenylacetic acid output was linearly related to grass but not to heather intake. The relationship between urinary phenylacetic acid output and grass intake could vary with different forages but that between orcinol output and heather intake was considered a useful index of heather intake.

7. Methods for the assay of urine phenols are discussed.

1. A pepsin + pancreatin method was used to assess the digestibility of pure cultures of rumen bacteria and mixed bacteria prepared from rumen fluid.

2. Individual species of Gram-negative rumen bacteria were highly digestible, whereas Gram-positive species, especially cocci, were more resistant to digestion.

3. A similar difference was observed microscopically with mixed rumen bacteria, but the influence of the relative proportions of Gram-positive and Gram-negative bacteria on the digestibility of bacterial protein in rumen fluid was small.