1. Thirty-three drugs known to stimulate the sympathetic nervous system have been screened for thermogenic properties. The results presented are for seven of them.

2. The drugs were tested in five animal models of obesity (genetic (mice and rats), hypothalamic (mice) and dietary (mice and rats)) as well asin leanmice. Energy-balance studies were undertaken using thecomparative-carcass technique as well as by measurement of daily oxygen consumption.

3. All seven drugs in obese animals tended to reduce body-weight and fat without loss of body protein: they acted by increasing metabolic rate without increasing food intake. They were much less effective in lean animals. These findings lend support to the concept that obesity is due to a diminished activity of the sympathetic nervous system.

4. Differences in the effectiveness of the drugs are discussed in relation to differences between the animal models of obesity. Ephedrine and tranylcypromine were found to be the most effective drugs in this series of experiments and a prima facie case is made for human clinical trials.

1. The possibility that viscous polysaccharides, such as guar gum, could lower post-prandial blood glucose levels in part by restricting carbohydrate solutions to a smaller area of small intestine was investigated in twenty healthy human volunteers.

2. Addition of guar gum (22.5 g/l) delayed the mouth-to-caecum transit time of a hypotonic lactulose drink, but did not affect gastric emptying.

3. When a 250 ml solution containing 50 g glucose was confined to a 550 mm length of intestine by an occluding balloon attached to an intestinal tube, maximum blood glucose response was significantly reduced (P < 0.05) though only by 0.9 mmol/l.

4. Addition of guar gum (36 g/l) had no effect on the distribution of a radio-labelled glucose drink (250 ml; 200 g glucose/l) in the small intestine, monitored using a gamma camera, although it significantly delayed gastric emptying (t½ (min): guar gum v. control 115 (SE 15) v. 73 (SE 8)).

5. Reduced contact area is unlikely to be one of the mechanisms by which guar gum improves glucose tolerance.

1. Pregnant rats were given control (46 mg iron/kg, 61 mg zinc/kg), low-Zn (6.9 mg Zn/kg) or low-Zn plus Fe (168 mg Fe/kg) diets from day 1 of pregnancy. The animals were allowed to give birth and parturition times recorded. Exactly 24 h after the end of parturition the pups were killed and analysed for water, fat, protein, Fe and Zn contents and the mothers' haemoglobin (Hb) and packed cell volume (PCV) were measured.

2. There were no differences in weight gain or food intakes throughout pregnancy. Parturition times were similar (mean time 123 (SE 15) min) and there were no differences in the number of pups born. Protein, water and fat contents of the pups were similar but the low-Zn Fe-supplemented group had higher pup Fe than the low-Zn unsupplemented group, and the control group had higher pup Zn than both the low-Zn groups. The low-Zn groups had a greater incidence of haemorrhaged or deformed pups, or both, than the controls.

3. Pregnant rats were given diets of adequate Zn level (40 mg/kg) but with varying Fe:Zn (0.8, 1.7, 2.9, 3.7). Zn retention from the diet was measured using 65Zn as an extrinsic label on days 3, 10 and 17 of pregnancy with a whole-body gamma-counter. A group of non-pregnant rats was also included as controls. The 65Zn content of mothers and pups was measured 24–48 h after birth and at 14, 21 and 24 d of age.

4. In all groups Zn retention was highest from the first meal, fell in the second meal and then rose in the third meal of the pregnant but not the non-pregnant rats. There were no differences between the groups given diets of varying Fe:Zn level. Approximately 25% of the 65Zn was transferred from the mothers to the pups by the time they were 48 h old, and a further 17% during the first 14 d of lactation. The pup 65Zn content did not significantly increase after the first 20 d of lactation but the maternal 65Zn level continued to fall gradually.

1. Three methods of measuring changes in body fat were compared in seventeen patients undergoing a 2-week course of intravenous nutrition. Patients received all nutrition intraveneously at a steady rate of infusion, calculated to supply energy at a rate equal to 1.5 times the resting metabolic expenditure measured before feeding. Fat change was estimated from measurements of skinfold thickness, by isotopic methods (neutron-activation analysis and dilution of tritiated water) and by daily analysis of expired gases.

2. The mean (with 1 SEM) gain in fat over the 2-week period was 1.14 (0.30) kg for skinfold measurement, 0.53 (0.62) kg for isotopic determination and 1.29 (0.22) kg for expired-gas analysis. There were no significant differences between the mean gains in fat measured by the three methods.

3. The results show that expired-gas analysis is the most sensitive technique, measuring change in fat content with an estimated precision of 0.26 kg. The isotopic method is less sensitive, with a precision of 2.38 kg, but provides a detailed description of body composition. In contrast to these highly-specialized techniques, both of which have limited application, the simple technique of measurement of skinfolds occupies an intermediate position of sensitivity, with a precision for measuring change in fat content of 0.85 kg, and also has the potential to measure total body fat content.

1. Rat pups were artificially reared (AR) from postnatal day 4 or 5 till day 20 or 21, by fitting them with gastric cannulas through which milk-substitutes could be infused automatically.

2. Three milk-substitutes were compared: milk M, the usual diet for AR studies, which was somewhat low in protein and very high in carbohydrate; milk A, which resembled rats' milk much more closely in composition; and milk isoM, which was based on the high-energy milk M but was made isoenergetic with milk A. Pups given these diets were termet ARM, ARA and ARisoM respectively. Siblings of the AR rats were left with their mothers to form a mother-reared (MR) control group. Rats were autopsied at 20 or 21 d.

3. Growth in body-weight of all groups of AR pups lagged behind that of their MR siblings for about the first week of AR, but the ARM group showed complete catch-up and the ARA group partial catch-up in body-weight during the second week. ARisoM rats were growth-retarded throughout.

4. Inspection of organ weights expressed relative to body-weight revealed disturbances of organ growth in all AR groups compared with MR animals. ARM rats showed excessive epididymal fat pad and liver weights, but deficits in gastrocnemius muscle, heart and adrenal weights. In contrast, ARA rats usually displayed increased spleen and stomach weights, but lower weight of interscapular brown adipose tissue. ARisoM rats had high brain, liver and stomach weights and low muscle and spleen weights relative to body-weight. All AR groups had elongated small intestines.

5. Hence the patterns of abnormal organ growth differed between groups. Those shown by the ARM and ARisoM groups seemed the more seriously abnormal. The diet approximating the composition of rats' milk (milk A) appears, as intended, to be an improved milk-substitute.

1. Casein was labelled with pairs of radioactive amino acids, lysine, tyrosine and leucine, one with I4C and the other with 3H, by jugular infusion into lactating goats followed by isolation of the double-labelled casein from the milk. Total milk protein was similarly labelled by jugular infusion of [35S]cystine. U-14C-labelled fraction- 1 leaf protein was isolated from lucerne (Medicago sativa) grown in an atmosphere of 14C02

2. The proteins were treated withdifferent levels(333 and667 mmol/kgprotein) offormaldehyde, glutaraldehyde and glyoxal.

3. Absorption from the small intestine was measured in sheep with fistulas in the abomasum and terminal ileum, using Cr-EDTA as the digesta flow marker, by introducing radioactive casein into the abomasum.

4. Lysine, tyrosine and cystine became increasingly unavailable for absorption from the small intestine of sheep with increasing levels of aldehyde. At the lower level (333 mmol/kg) the proportions of the amino acids that were unavailable were 0.192, 0.051 and 0.123 respectively. At the higher level of formaldehyde (667 mmol/kg) the corresponding values were 0.335, 0.201 and 0.432 respectively. Leucine was not made unavailable with formaldehyde.

5. The proportions of lysine, tyrosine and leucine that were unavailable were higher, on a molar basis, after treatment of the proteins with the dialdehydes glutaraldehyde and glyoxal than after treatment with formaldehyde. However, the extent of protein protection provided by the dialdehydes in the rumen, measured using an in vitro procedure, was lower.

1. Three steers, each fitted with a rumen cannula and a re-entrant cannula in the proximal duodenum, were offered diets consisting of a barley-based concentrate and chopped hay at a daily intake of 61 g/kg live weight0-75 given in three. equal meals. The ratio, concentrate: hay was changed from 50: 50 to 90: 10 and then to 100: O in successive periods of 12–18 weeks and the flow and composition of digesta at the duodenum was measured over 48-h periods on each dietary treatment.

2. Samples of bacteria and protozoa were separated from rumen contents and the proportions of bacterial and protozoal nitrogen (N) in duodenal digesta were estimated using 2, ddiaminopimelic acid (DAPA) and 2- aminoethylphosphonic acid (AEP) as markers. On separate occasions, radioactive sulphur (35S) was infused into the rumen for 48 h and digesta collected over the final 24 h; the specific radioactivity of S in microbial and digesta fractions was used to estimate the proportions of microbial N.

3. 35S gave reproducible and apparently reliable estimates of microbial protein formation: the proportion of microbial N in digesta was significantly higher (P < 0.05) for the 50:50 diet than for the other treatments but the energetic efficiency of microbial protein formation did not differ significantly between diets.

4. Estimatesof bacterial N based on DAPA concentrations were highly variable and frequently impossibly high. It is suggested that many of the anomalous values were the result of non-representative sampling of the rumen microbial population and that this is particularly likely to occur when conditions within the rumen are unstable. AEP was found to be unsuitable as a marker for rumen protozoa as considerable concentrations of this substance were found also in rumen bacteria.

1. Twelve steers fitted with rumen cannulas were used in three separate experiments to investigate the effects of the presence or absence of rumen ciliate protozoa on methane production. The diet consisted of 850 g barley and 150 g protein supplement/kg, and was given in three feeds daily at a restricted level of 61 g/kg live weight0.75. Animals were defaunated initially by allowing ad lib. consumption of this diet and were then maintained ciliate-free by isolation or were faunated by inoculation with a mixed ciliate suspension. Samples of rumen fluid were taken routinely for the assessment of microbial populations and for volatile fatty acid (VFA) analysis and energy and nitrogen balances and digestibility measurements were made at intervals while animals were confined in respiration chambers.

2. In each experiment the rumen VFA proportions changed from a high-propionate pattern under ciliate-free conditions to a low-propionate, high-butyrate pattern in the presence of ciliates: differences between treatments were highly significant (P < 0.001). There were also marked differences between treatments in CH, production but a reliable comparison was possible only in Expt 3, in which CH4 was significantly higher (P < 0.001) in the presence of a rumen ciliate population. In Expt 3 the increased loss of energy as CH4 in the faunated animals amounted to 44 MJ/100 MJ energy intake.

3. Stoichiometric estimates of CH4 production derived from the observed VFA proportions showed good agreement with CH4 production as measured in respiration chambers. On average, the stoichiometric CH4 values overestimated CH4 production by a factor of 1.08.

4. Highly significant linear relationships (P < 0.001) were observed between the molar proportion of each major VFA and the quantity of CH4 produced: the proportion of propionic acid was inversely related to CH4 and showed the lowest residual standard deviation of all the relationships examined.

5. The losses of energy in faeces and urine did not differ between treatments hence the increased loss of energy as CH4 in the faunated animals resulted in a significant reduction in the metabolizability of the diet from 0.73 to 0.69 (P < 0.05). No significant differences were detected between treatments in heat production, apparent digestibility coefficients or N balance.

6. It is suggested that the rumen ciliates, by modifying the rumen VFA proportions, are directly responsible for the increased CH4 production in faunated animals.

1. The metabolism of glucose has been studied in 48-h-old unanesthetized fasting and post-absorptive sucking piglets.

2. Both [6-3H]- and [U-14CJglucose were administered either by a single injection method or by a primed infusion technique. The rates of glucose turnover and recycling were estimated under steady-state conditions.

3. The rates of glucose turnover and recycling in 48-h-old fasting or post-absorptive piglets were not statistically different when measured using the single injection technique or the printed infusion method.

4. The mean (with SE) rate of glucose turnover was 65.8 (2.5) in post-absorptive and 31.1 (1–9), μ mol/kg per min in fasted newborn pigs. Glucose utilization was linearly related to blood glucose concentration; regression analysis indicated a y-intercept of 7.2, μ mol/kg per min.

5. As tested by arterio-portal differences the gut was not releasing glucose or galactose in 5 h-post-absorptive sucking newborn pigs. Thus, the higher rates of glucose turnover in post-absorptive newborn pigs compared with fasting ones suggest that hepatic glucose production is enhanced in post-ahsorptive sucking piglets.

6. The mean (with SE) rates of glucose recycling were four times higher in post-absorptive piglets than in tasting ones, i.e. 14.4 (1.6) and 3.7 (0.5)% of [6-3H] glucose turnover respectively. As liver glycogen was exhausted in 48-h-old sucking piglets, this suggests that hepatic glucose production results from gluconeogenesis.

1. In two separate trials male and female Wistar rats, 12 weeks of age, were either killed as a preliminary control group, ad lib.-fed or undernourished for 4 weeks until one-third of their 12-week body-weight was lost.

2. Food intakes, urinary and faecal collections and measurements of standard metabolic rate were made at one-weekly intervals on both the ad lib.-fed and undernourished animals of both sexes.

3. The bodies of the preliminary controls, the ad lib.-fed and the undernourished animals of both sexes were analysed for protein and fat, and the weights of four fat depots, two muscles and the major organs of all groups were determined.

4. Measurements of lipid synthesis rate (LSR) and lipoprotein lipase (EC 3.1.1.34) (LPL) activity in the four fat depots and measurements of whole-body protein synthesis rates were carried out on animals of both sexes in each group.

5. Although both sexes lost the same proportion of body-weight the females required more food on a body-weight basis than the males during the undernutrition period. The females absorbed significantly more energy on a body-weight basis during undernutrition and so were less efficient than the males at withstanding nutritional stress.

6. There were no significant differences between males and females, on a body-weight basis, in the excretion of nitrogenous waste products (urinary nitrogen, creatinine, hydroxyproline or NT-methylhystidine) suggesting that there were no differences between the sexes in protein sparing during undernutrition.

7. There were no differences between males and females in the proportions of body fat and protein used during the period of undernutrition or in the sites of the body from which the protein and fat were mobilized.

8. There were no differences between males and females in the way they responded to undernutrition by altering LSR, LPL activity or whole-body protein synthesis rates. Both undernourished males and undernourished females maintained synthesis of lipid, on a per g tissue basis and whole-body protein synthesis at the level found in well-nourished animals of the same sex.

1. Female Wistar rats (5 and 11 weeks old) were either left intact or ovariectomized. Animals of each age- and treatment-group were either ad lib.-fed or undernourished for 4 weeks.

2. The bodies of all animals were analysed for protein and fat and the weights, lipid synthesis rate and lipoprotein lipase (EC 3.1. 1.34) activity of four fat depots were determined.

3. The well-nourished ovariectomized animals of both age-groups gained weight more rapidly than the well-nourished intact animals of the same age, but there was no effect of ovariectomy on body composition, lipid synthesis rate or lipoprotein lipase activity in either the well-nourished or the undernourished animals of either age-group.

4. There was a greater efficiency of energy utilization found in the ovariectomized animals than in the intact animals regardless of age or nutritional status.

1. Four mature Friesian cows were used in a Latin square design experiment to measure the effects of abomasal casein infusion on milk production and the kinetics of glucose, acetate and palmitate in jugular venous blood.

2. The basal ration contained 110 g crude protein/kg dry matter and was offered at the rate of 4 kg hay, 3 kg alkali-treated straw cubes plus 9 kg concentrates/d. Treatments were infusion of casein into the abomasum at the rate of 0, 240 or 460 g/d. The experiment started 1 week post partum. Experimental periods lasted 21 d.

3. [6-3H]glucose, [9, 10-3H]palmitate and [U-14C]acetate were continuously infused into one jugular vein for measurements of whole-body flux rates.

4. The yield of milk protein was increased significantly (P < 0.05) by the first increment of casein infusion, with no further increases.

5. Casein infusion had no effect on blood concentrations of glucose, acetate, palmitate or growth hormone. Mean flux rate of glucose increased with each increment of casein. The effect at 460 g/d was higher (P < 0.1) than with no infusion. Flux rates of palmitate and acetate were increased (P< 0.1) by the first increment of casein. The proportion of blood bicarbonate derived from acetate increased (P < 0.05) with the first increment of casein.

6. Protein status of cows early in lactation influences the dynamics of glucose and fatty acid metabolism.

1. The effects of vitamin D3 (D3) on serum levels of 1, 25-dihydroxyvitamin D3 (1, 25(OH)2D3), ionic calcium, total Ca and phosphorus in chicks were studied from the time of hatching until sexual maturity.

2. Chicks fed on a diet low in D3 showed a serum level of 1, 25(OH)2D3 higher than that in chicks on a normal-D3 diet, for both sexes and at any given age.

3. A dramatic increase in the serum level of 1, 25(OH)2D3 occurred in female birds approaching sexual maturity and in laying hens raised on the low-D3 diet the level was five times that of their counterparts raised on a normal-D3 diet.

4. Theserum 1, 25(OH)2D3levelin adultmalesin thelow-D3groups wasseven timesthatofthoseon thenormal-D3 diet.

5. The serum level of 25-hydroxyvitamin D3 remained relatively unchanged at weeks 2 and 15 in birds on a low D3 intake as well as in those fed on a normal-D3 diet. Nevertheless, the levels of 25-hydroxyvitamin D3 were different between the two groups.

6. No significant change was observed in the level of ionized serum Ca in relation to dietary regimen, but there was an increase in total Ca concentration in females with the onset of reproduction.

7. The serum P level decreased gradually with age, reaching a minimum value 3 and 8 weeks before laying commenced in the groups on low- and normal-D3 diets respectively. An increase was observed when the hens began laying.

8. Chicks adapted to a low-D3 diet by elevation of their plasma level of 1, 25(OH)2D3. The mechanism by which this is achieved is not known, but the results suggest that parathyroid hormone, Ca and P are unlikely to play roles in the adaptive increase in the level of 1, 25(OH)2D3 in the blood of chicks given a minimal amount of D3. The possibility that the rate of degradation of 1, 25(OH)2D3 is greatly reduced under these conditions cannot be excluded and this could account for the level of this metabolite in those birds.

1. Investigations were carried out to establish the total biopotency of the natural vitamin E isomers in barley compared with that of DL-α-tocopheryl acetate.

2. The chick was used as an experimental animal. Prevention of nutritional encephalomalacia (NE) and chick liver-storage and plasma-storage assays of vitamin E were the methods used in the study. The individual tocopherols and tocotrienols, both in the tissue samples and in the grain and barley oil, were analysed using high-pressure liquid chromatography (HPLC) with fluorescence detection. The diagnosis of NE was based on careful clinical and histopathological observations.

3. It can be concluded from the results that full protection against NE in the chicks was obtained with a supplementation level of 7.5 mg DL-α-tocopheryl acetate/kg diet (i.e. a total vitamin E content of 11.20 mg/kg diet) or with a supplement of 8.7 g barley oil/kg diet (i.e. a total vitamin E content of 22.99 mg from barley oil/kg diet). This gave a biopotency factor of 0.49 for barley for prevention of NE of the chicks, as compared to that of DL-α-tocopheryl acetate.

4. Using regression analysis a statistically linear relationship could be observed between the total dietary vitamin E level and the response, as measured by the total vitamin E content in the liver and plasma, both in the groups supplemented with DL-a-tocopheryl acetate and in the groups supplemented with corresponding amounts of vitamin E in barley oil. The liver and plasma responses to the total vitamin E in the barley-oil diet compared with those of the DL-a-tocopheryl acetate reference diet gave identical values for the regression coefficients, i.e. in both liver-storage and plasma-storage assays the value for slopes of dose-response lines was 0.37. This means that the biopotency of the total vitamin E in barley was 37% of that of dietary DL-a-tocopheryl acetate. Thus, barley is not as rich a source of vitamin E as could be supposed on the basis of the chemical determination of its total vitamin E content.

5. It was possible to verify this experimentally established biopotency of 0.37 for the total vitamin E in barley by converting the chemically determined amounts of the vitamin E isomers in barley into DL-α-tocopheryl acetate equivalents by multiplying them with internationally accepted potency factors for the individual natural isomers (DL-α-tocopheryl acetate 1.00, D-α-tocopherol 1.49, D-β-tocopherol 0.60, D-gamma;-tocopherol 0.1 5, D-α-tocotrienol 0.37).

6. In spite of the high proportion of α- and β-tocotrienols in the barley-oil diets (about 60% of the total vitamin E content), only traces of these isomers could be detected in the plasma and none could be detected in the liver. On the other hand, calculation of the individual hiopotencies for the different isomers in the barley-oil diet by comparing the dose responses, diet: liver, separately for each isomer with those of DL-α-tocopheryl acetate, resulted in biopotency values for α- and β-tocopherol which were twice as high as the internationally accepted conversion factors. These results of the present study tempted the authors to draw the conclusion that there may have been a chemical reduction of the α- and β-tocotrienols to the corresponding tocopherols before entering the liver.

1. Sheep offered a roughage diet for 4 h daily were injected intravenously with glucose before and at various times after feeding. The insulin secretory response to glucose and the rate of disappearance of injected glucose were determined.

2. While the basal concentration of plasma insulin was unchanged, the base-line plasma glucose concentration tended to decrease during the meal.

3. The glucose load brought about an increase in the plasma insulin concentration at each injection, but the insulin response to glucose and the rate of glucose disposal were increased during the meal.

4. On varying the time of feeding between 08.00 and 16.00 hours, the increase in the insulin response to glucose and in the rate of glucose disposal always appeared to be related to the giving of food, independent of the time food was offered.

5. It is concluded that feeding increases the insulin response to an intravenous glucose load even when the increase in the basal level of plasma insulin on feeding is very modest in sheep given a roughage diet. The increased insulin response and glucose disposal rate following feeding did not appear to be related to diurnal rhythms in insulin secretory activity or glucose metabolism.

1. Whole-body and tissue fractional protein synthesis rates were determined in chronically-catheterized ovine fetuses at 120–130 d of gestation following an 8 h continuous infusion of L-[U-14C]-or L-[2, 3, 5, 6-3H]tyrosine.

2. From the net utilization of tyrosine by the fetus, corrected for apparent oxidation, and tyrosine concentration in the fetal carcass protein, whole-body protein synthesis was estimated to be 63 g/d per kg.

3. Following 8 h of infusion of labelled tyrosine the ewes were killed and fetal tissues were removed for the determination of tyrosine specific activity. The fractional rate of protein synthesis (k3) was calculated from the specific activity ratio, protein bound: intracellular free tyrosine. Tissue k, values for the liver, kidney, lungs, brain, skeletal muscle and small intestine were 78, 45, 65, 37, 26 and 93% /d respectively.

4. The absolute rate of synthesis was calculated by multiplying the tissue protein content by k2. Muscles, gastrointestinal tract, liver and lungs contributed approximately 20.5, 20.5, 14.4 and 9.4% respectively to whole- body protein synthesis.

5. The efficiency of protein synthesis as expressed by the RNA activity was higher in liver, lung and brain followed by kidney, skeletal and cardiac muscle.

1. Immature male Wistar rats were given a zinc-depleted semi-synthetic diet (2 mg Zn/kg diet) for 4 weeks. Control groups received a similar diet containing 60 mg Zn/kg diet, either ad lib. or in amounts matched to the consumption of the Zn-deficient group. Rates of growth and food conversion efficiency were markedly lower in the Zn-deficient group compared to controls.

2. The phloridzin-sensitive uptake of galactose and 3-0-methyl glucose by everted jejunal rings from each group was measured over a range of concentrations. Kinetic analysis revealed that the maximum transport rate (Vmax)for the uptake of both sugars was significantly higher in the Zn-deficient group than in the control groups.

3. The villi, microdissected from samples of proximal jejunum, were significantly shorter and narrower across the base in the Zn-deficient group than those of the control groups, but were present in greater numbers per unit serosal area.

4. The present study shows that Zn deficiency leads to significant morphological changes in rat small intestine, which are accompanied by an increase in the capacity of the tissue to take up sugars across the mucosal membrane.

1. Sheep were fed once daily with 300 or 600 g dried perennial ryegrass (Lolium perenne) or white clover (Trifolium repens). Total rumen contents were sampled once daily at various intervals up to 24 h after feeding for 6 d.

2. Total weight of fresh rumen contents and dry matter (DM), organic matter and particulate matter concentrations were measured and the particle size distribution of rumen contents was analysed using a wet sieving technique.

3. There was a similar exponential loss of DM from the rumen with time after feeding for both perennial ryegrass and white clover. The rate of DM disappearance during the first 3 h after feeding was ten times greater for the higher intake and seven times greater for the lower intake than during the subsequent 21 h. The percentage of particulate material disappearingduring the first 3 h was greater for white clover (57% )than for perennial ryegrass (49%).

4. Particle size distribution analysis of rumen contents showed that the most rapid rate of particle size reduction occurred during the first 3 h after feeding. Despite a rapid rate of breakdown there was no increase in the total weight of small particles in the rumen.

5. The rate of reduction in particle size during this period was greater for white clover than perennial ryegrass.

6. Microscopic examination of rumen particles showed a difference in particle geometry between white clover and perennial ryegrass which appeared to be related to anatomical features of the plant.

1. The primary growth of lucerne (Medicago sativa) was ensiled after treatment with either formic acid alone (4.1 litres/t; silage F) or with formic acid and either formaldehyde (30.5 g/kg crude protein (nitrogen x 6.25; CP); silage FF), glutaraldehyde (44.2 g/kg CP; silage FG) or a mixture of the two aldehydes at approximately half their individual application rates (silage FFG).

2. Compared with formic acid alone, both formaldehyde and glutaraldehyde reduced protein breakdown and carbohydrate fermentation during ensiling. The extent of protein protection afforded within the silo was similar for the two aldehydes, whereas formaldehyde was more effective in restricting carbohydrate fermentation. The effect of treatment FFG on silage fermentation was confounded by the silo bag bursting and the development of a clostridial-type fermentation. All aldehyde treatments reduced silage soluble-N content but N disappearance when the silages were incubated in polyester bags in the rumen was high for all silages and reductions due to the aldehydes were small.

3. Silage digestion was studied in four mature sheep each fitted with a rumen cannula and re-entrant cannulas in the proximal duodenum and distal ileum.

4. The apparent digestibility of organic matter (OM) in the whole tract was reduced (P < 0.05) to a similar extent by both aldehydes, whereas rumen OM digestion was reduced (P < 0.05) more by glutaraldehyde than by formaldehyde. The effects on digestion appeared to be due to the action of the aldehydes on the foods rather than to any adverse influences of the aldehydes on the metabolism of the rumen microbes because, although rumen ammonia levels were lower (P < 0.05) when the aldehyde-treated silages were given, rumen casein-degrading activity, the degradation of different feedstuffs when incubated in polyester bags in the rumen and microbial N flow at the duodenum did not differ (P < 0.05) between silages.

5. All aldehyde treatments decreased (P < 0.05) the apparent digestibility of N in the whole tract. Silage N degradability in the rumen was also decreased (P < 0.05) from 0.82 for silage F to 0.67, 0.60 and 0.62 for silages FF, FG and FFG respectively, and consequently non-ammonia-N (NAN) flow at the duodenum increased (P < 0.05). The aldehydes did not adversely affect the apparent digestibility of NAN in the small intestine, and net NAN absorption from the small intestine increased from 8.8 g/d with silage F to 11.4, 15.3 and 14.2 g/d with silages FF, FG and FFG respectively. Both the decrease in N degradability in the rumen and the increase in net NAN absorption from the small intestine were greater (P < 0.05) with glutaraldehyde than with formaldehyde.