MAGNESIUM ABSORPTION FROM THE DIGESTIVE TRACT OF SHEEP

Transcription

1 Quaterly Journal of Experimental Physiology (1984), 69, Printed in Great Britain MAGNESIUM ABSORPTION FROM THE DIGESTIVE TRACT OF SHEEP Department of Animal Physiology and Nutrition, University of Leeds, Leeds LS2 9JT SUMMARY Factors affecting absorption of Mg from the ovine rumen have been studied using either a pouch constructed from part of the dorsal rumen or by an isolated washed rumen technique in vivo. Net absorption of Mg against the prevailing electrochemical gradient was observed. An increase in the K/Na ratio within the rumen led to an increase in the potential difference across the rumen wall, blood positive, and to a decrease in the net efflux of Mg from the rumen. This decrease was due to an increase in Mg influx into the rumen. The addition of ammonium chloride (30 mmol/l) to the rumen contents also led to a reduction in net Mg absorption but to no significant change in potential difference. The effects of high K/Na ratio and high ammonium ion concentration within the rumen were additive in causing decreases in net effluxes of both Mg and Na. An inverse relationship was demonstrated between the Ca concentration in the rumen and the net absorption rate of Mg. It was concluded that the efflux ofmg across the rumen wall depends at least in part on a functional system for Na transport. INTRODUCTION Although there is evidence for a relatively small degree of net absorption of Mg from both the small intestine (Care & van't Klooster, 1965; Phillipson & Storry, 1965) and the large intestine (Grace, Ulyatt & MacRae, 1974), it is now believed that the major site of net absorption of Mg in sheep is from the reticulo-rumen (Pfeffer, Thompson & Armstrong, 1970; Tomas & Potter, 1976). Some may also be absorbed from the omasum as has been shown in cattle by Edrise & Smith (1979). The absorption of Mg takes place not only by diffusion but also by the active transport of Mg ions across the rumen epithelium (Martens, Harmeyer & Michael, 1978; Brown, Care & Pickard, 1978); this active transport is relatively more important at lower concentrations of Mg within the rumen. Changes in the relationship between passive transport of Mg and intraruminal Mg concentration can be effected by alteration in the potential difference (p.d.) across the rumen epithelium. For example, this p.d. can be increased by increasing the ratio of K/Na within the rumen (Scott, 1966). In studies involving an isolated washed rumen in an anaesthetized sheep, the p.d. was found to vary linearly with the logarithm of the K ion concentration at a constant concentration of Na ions (Ferreira, Harrison, Keynes & Nauss, 1966). In the absence of Na ions or in the presence of 10-4 mol ouabain/l there is a drastic fall not only in p.d. but also in short circuit current across an isolated sheet of rumen epithelium in vitro, thus indicating their dependence on the active transport of Na ions (Ferreira, Harrison & Keynes, 1966). The relationship between net transport of Mg and that of Na was shown by Care, Farrar & Pickard (1982).

2 578 Surgical techniques METHODS Rumenpouch. A rumen pouch was established under halothane/02 anaesthesia as a chronic preparation in the dorsal rumen of eighteen sheep and in the ventral rumen of one sheep (Komarek, Leffel, Brown & Mason, 1959; Foss & Black, 1972). Twelve of these preparations survived for at least 1 month, the longest being 13 months. The principal reason for failure was the development of a small fistula in the septum dividing the pouch from the body of the rumen. A small nylon or latex cannula was placed in the wall of the pouch and brought to the surface via a stab wound in the body wall. The volume of pouch so formed was ml. During the first week after surgery warm 0 14 mol NaCl/l was used to flush out the pouch daily. Subsequently, the pouch was filled each day with filtered rumen contents from a sheep fed the same diet as the experimental animal. After the addition of this fluid to the pouch, the cannula was left with a plastic tube in its bung arranged in such a way as to allow the release of gaseous products of fermentation but retain the liquid contents of the pouch. Histological examination of the pouch in sheep D500, 13 months after its preparation, showed that the height of the papillae was reduced in comparison to the adjacent dorsal ruminal epithelium and that keratinization was increased. Basal layers included some eosinophilic nuclei, a feature absent in the intact ruminal epithelium. During the experimental period regular contractions of the pouch were observed, a necessary criterion for normal function. Washed reticulo-rumen. A large rumen fistula was prepared in the dorsal sac of a large sheep by first removing a circle of skin about 5 cm in diameter and parting the underlying layers of muscle and peritoneum to reveal the dorsal sac of the rumen. The rumen wall, muscle layers and peritoneum were then sutured together around the periphery of the skin incision before the rumen was incised to produce the fistula. This fistula was normally closed with a removable foam rubber bung secured with inner and outer flanges held in position by tape ties (Martens & Rayssiguier, 1980). Measurements of Mg absorption (a) Rumen pouch. Residual rumen contents were removed from the pouch by gentle suction coupled with pressure to the abdomen ventral to the cannula. The pouch was washed repeatedly with warm 0 14 mol NaCl/l until the effluent showed no colour. The pouch was then filled with 0-14 mol NaCl/l at 37 C and left for 20 min before emptying. This procedure was repeated for a further period of 20 min before the pouch was filled with unlabelled experimental solution (Table I) and left to equilibrate for 15 min. The pouch was then emptied and filled with a known volume of experimental solution labelled with 51Cr EDTA (Amersham International PLC, Bucks). After 5 min a 10 ml sample was taken by syringe via the tube in the cannula bung. This sample was deemed to be taken at time zero: absorption of Mg was measured over the next 2 h, after which the pouch was emptied and a final sample taken. The pouch was then flushed at least five times with the next unlabelled experimental solution.

3 MAGNESIUM ABSORPTION IN SHEEP 579 Table 1. Composition of solutions Composition of washing buffer (mmol/l) Composition of solution used with rumen pouches (mmol/l) NaCl KCl NaHCO3 25 KHCO3 20 Na2HPO4 * 12H20 2 Glucose 5 Na acetate 30 Propionic acid* 10 Butyric acid* 5 CaC12.6H20* 2 4 MgC12. 6H20* * A solution of these substances should be made together before adding to a solution of the others, so as to avoid precipitation. The absorption rate of Mg, A (mmol/2 h) was calculated as follows: A V C [5'Cr ] = where C1 = Mg concentration (mmol/l) in pouch fluid at time = 0, C2 = Mg concentration (mmol/l) in pouch fluid at time = 2 h, 5Crl = 51Cr concentration (uci/l) in pouch fluid at time = 0, 5'Cr = 51Cr concentration (4uCi/l) in pouch fluid at time = 2 h, V = initial volume of experimental solution (usually 0 051). When 28Mg (Institute of Nuclear Chemistry, Julich, F.R.G.) was present in the experimental solution (0-25,uCi/l), [14C]polyethylene glycol 4000 (Amersham International PLC, Bucks) was used as fluid marker at a concentration of 4,uCi/l. (b) Washed reticulo-rumen. The sheep was trained to stand quietly, restrained merely by a loose neck-yoke. The fistula plug was removed and the rumen contents sucked into a plastic bottle using a vacuum pump. The contents were then stored at 38 C until the end of the experiment when they were returned to the sheep. Two litres washing buffer (Table 1) at 38 C was poured into the rumen and agitated manually in order to dislodge material from the ruminal epithelium. The buffer was removed by suction and the procedure repeated until the effluent was free from rumen contents. The washed rumen was isolated according to Martens & Rayssiguier (1980). The base of the oesophagus was temporarily occluded by an inflated cuff and the saliva collecting cranial to the obstruction was pumped into the omasum via a Foley catheter inflated in the reticulo-omasal orifice. Air was allowed to bleed into the lower oesophagus when suction was applied in order to prevent the oesophagus from collapsing and occluding the outflow of saliva. The experimental solutions, based upon the composition of the washing buffer, were introduced into the washed reticulo-rumen and continually gassed with carbon dioxide via a tube fitted with a diffusing nozzle. During an experiment the fistula in the rumen was sealed by inner and outer foam rubber flanges held together on either side of a bung by tapes tied externally. Tubes for saliva

4 580 collection and recirculation, carbon dioxide and air bleed to the saliva collector ran through both flanges and bung. Before each measurement of Mg absorption, 2 1 of the experimental solution without its fluid marker (Cr EDTA or polyethylene glycol (4000 PEG)), warmed to 38 C, was added to the reticulo-rumen via a plastic tube inserted between the inner flange of the fistula plug and the abdominal wall. After 30 min this solution was removed by suction. Two litres of experimental solution containing either PEG (2 5 g/l) or Cr EDTA (Downes & McDonald, 1964) at 38 C was infused rapidly into the rumen and the plastic tube used for sampling arranged so that its tip lay in the ventral sac of the rumen. After 10 min a 10 ml sample was taken which was deemed to be taken at time zero. Further samples were taken at 20, 40 and 60 min later. The calculation of Mg absorption rate was similar to that used with the rumen pouch. Measurement ofpotential difference across the rumen wall The method used was that of Dobson & Phillipson (1958) except that the blood was not allowed to clot in the venous catheter but was continually mixed with heparinized saline, thus facilitating repeated observations. Analytical methods Mg concentrations were measured by atomic absorption spectrophometry (Unicam SPI91; Pye, Cambridge) using 010% lanthanum chloride as diluent. Ca concentrations were measured either by atomic absorption spectrophotometry (Pye Unicam SPI91) or by an Autoanalyser technique (Technicon Instruments, Basingstoke; method N 36b). Na and K concentrations were measured by emission spectroscopy (Unicam SP90; Pye, Cambridge). Concentrations of Cr (Cr EDTA) were measured by atomic absorption spectrophotometry. PEG (4000) was measured by the turbidimetric method of Hyden (1955). Osmotic pressure was measured by depression of the freezing point (Osmette, Precision System Inc, Newton, MA, U.S.A.). Measurement of radio-isotopes All samples were counted for a sufficient time to ensure counts, thus reducing counting errors to the region of + 1 %. 28Mg and 51Cr were counted in a well-type scintillation counter using a Nal crystal (Gamma Guard 150, Tracer Laboratories, Weybridge, Surrey). Measurement of 14C in the presence of 28Mg was delayed for 2 weeks to allow complete decay of the 28Mg. 14C was counted in a liquid scintillation counter (Tracer Laboratories, Weybridge, Surrey) after 1 ml pouch fluid had been dispersed in 10 ml scintillant (Bray, 1960). RESULTS The relationship between intraruminal Mg concentration and rate of Mg absorption Two sheep with a rumen pouch were used in this study. The duration of each absorption period was 2 h and the compositions of the experimental solutions used were (in mmol/l): NaCl, 90; KCI, 30; CaCl2, 4; and MgCl2, Mannitol was used to maintain a constant initial osmotic pressure (290 mosmol/kg).

5 MAGNESIUM ABSORPTION IN SHEEP = 40- o 30- E '/ mmol Mg/l Fig. 1. Effect of increasing the concentration of Mg in the fluid within a rumen pouch on the rate of net efflux of Mg from the pouch. Vertical bars represent S.E.M. n = 5 at each Mg concentration. *, sheep D573; E), sheep D500. The effect of increasing the concentration of Mg within the rumen pouch on the net efflux of Mg is shown in Fig. 1. A curvilinear positive relationship was obtained over the entire range of Mg concentration studied. Using the Eadie-Hofstee plot of the ratio of Mg efflux rate to Mg concentration in the ruminal fluid as abcissa, against the rate of Mg efflux as ordinate (Martens, 1983), the Michaelis-Menten constant (Kin) was found to be 7-5 mmol/l and the maximum rate of efflux of magnesium from the rumen pouch (Vmax) was 1-6,amol/min. The regression coefficient of this linear relationship was The effect of osmotic gradient across the rumen wall of a rumen pouch on the efflux of Mg ions The composition of the normal experimental solution used was (mmol/l): NaCl, 80; KCI, 20; MgCl2, 3; and CaCl2, 4; this had an osmotic pressure of 204 mosmol/kg. Mannitol was added to portions of this solution to produce osmotic pressures of 224, 298 and 374 mosmol/kg, respectively. In each of two sheep using solutions which were approximately isotonic with plasma, a 99%O recovery of added 51Cr EDTA was obtained, after correction for the small volume of experimental solution remaining in the rumen pouch at the end of the absorption period. No estimates of recovery of 5'Cr EDTA were made from the hypotonic solutions which were only used in the one study. The rate of effilux of Mg from each of these fluids in the rumen pouch was compared in four experiments. There was a significant (P < 0-001) negative correlation between the net rate of efflux of water (y), measured by the use of a tracer addition of 51Cr EDTA to the fluid, and the initial osmotic pressure of the pouch fluid (x). From the regression equation, y = x linking these two parameters, it can be seen that net flux of water was zero at an osmotic pressure of pouch fluid of 324 mosmol/kg, i.e. fluid hyperosmotic relative to plasma of this sheep (291 mosmol/kg). This takes no account of the small loss of 5'Cr EDTA which has been reported to occur from the rumen when

6 a725. f1)low Na/high K -i 0 2) Intraruminal Ca concentration (mmol/l) Fig. 2. The effect of intraruminal concentration of Ca on the net rate of effiux of Mg from a rumen pouch in three sheep. For explanation, see text. Figures in parentheses represent the number of observations. hypotonic fluids are used (Dobson, Sellers & Gatewood, 1976). The net rate of efflux of Mg was independent of the osmotic pressure of the pouch fluid over the range employed, although the net rate of efflux of K and Na were both positively correlated (P < 0 001) with rate of net water efflux. The influence of intraruminal Ca concentration upon the net efflux of Mg Four experiments were conducted using a rumen pouch in two different sheep. The pouch fluid initially contained mmol Na/l, mmol K/l and mmol Mg/l and had an initial volume of 60 ml. There was an over-all reduction in the rate of efflux of Mg in association with an increase in the Ca concentration within the pouch (Fig. 2). In two studies with a third sheep, one carried out with a pouch fluid containing 120 mmol Na/l and 20 mmol K/1 (high Na/low K) and the other with K/Na ratio of 120/20 (low Na/high K), an increase in the Ca concentration within the pouch fluid from 2 to 8 mmol/l was associated with a decrease in the net efflux of Mg from (11) to (12),tmol/3 h and (13) to (10),umol/3 h, respectively. Both solutions contained 2 5 mmol Mg/l. There was no significant effect of intraruminal Ca concentration on the p.d. across the wall of the rumen pouch but net Ca absorption was significantly increased (P < 0 001) in both sheep studied. In one experiment, 28Mg was added to the pouch fluid so that the effect of ruminal Ca concentration on the one-way absorption of Mg could be assessed. The absorption rate of 28Mg was depressed (P < 0-001) by increasing the intraruminal Ca concentration whereas there was no significant effect of this on the influx of Mg into the pouch fluid (Fig. 3).

7 MAGNESIUM ABSORPTION IN SHEEP E (3) (4) (3) Mean Ca concentration (mmol/l) Fig. 3. The effect of varying the concentration of Ca upon the rate of efflux of 28Mg and 24Mg from a rumen pouch. Total bar height represents the mean (± S.E.M.) one-way 28Mg efflux; the hatched portions represent the net 24Mg efflux. The open portion thus represents the one-way influx of Mg. Figures in parentheses represent the number of observations at each Ca concentration. The role of Na and K concentration in the absorption of Mg from the rumen Rumen pouch. Seven sheep were used in this study. Using sheep D500 and D573, the effect on Mg efflux was studied of increasing the concentration of KCl in a solution of the following composition (mmol/l): MgCl2, 2; CaCl2, 4; NaCl, 30; KCl, and mannitol to maintain equality of osmotic pressure. Increasing the concentration of KCI caused a significant (P < 0-001) reduction in the net efflux of Mg from the pouch fluid. The regression equations obtained between Mg efflux (y,tmol/2 h) and K concentration (x mmol/l)werey = x; n = 20, r = 0-92 andy = x; n = 20, r = 0-84 for sheep D500 and D573, respectively. In the remaining studies the sum of the Na and K concentrations was kept constant throughout one series of experiments, so as to maintain a constant level of chloride which resembles the normal situation (Horn & Smith, 1978). Thus, with sheep D500, a high-k solution contained 20 mmol NaCl/l and 80 mmol KCl/l whereas a low-k solution contained 80 mmol NaCl/l and 20 mmol KCl/l. The concentration of Ca was 4 mmol/l and the Mg concentration in these solutions was varied in order to study the effect of K/Na ratio on the relationship between Mg effiux and Mg concentration within the rumen pouch. Chloride was the sole anion present. The results are shown in Fig. 4. The shape of the curve is similar to that shown in Fig. 1 and it can be seen that the net efflux of Mg, at a given Mg concentration, was consistently lower from solutions with a high K/Na ratio. Although there was no significant effect of Mg concentration on the p.d. across the wall of the pouch, this

8 E/ 0S 60/ a) 20 2 ' ' l Mean Mg concentration (mmol/l) Fig. 4. The effect of varying the initial K/Na ratio in a rumen pouch from 20/80 (V-V) to 80/20 ( ) on the rate of net efflux of Mg from the pouch at differing Mg concentrations within the pouch field. p.d. was increased from mv (S.E.M.) (n = 5) to (5) mv by changing from a solution with a low (20/80) K/Na ratio to one with a high (80/20) K/Na ratio. The inclusion of 28Mg within the pouch fluid of sheep DR19 containing 2 mmol Mg/l enabled measurements to be made of both influx and effilux of Mg from the pouch fluid. There was again a significant (P < 0 01) reduction in the net efflux of Mg from (4) to 1 1 +± 1 (4),umol/ I 5 h caused by raising the K/Na ratio within the pouch from 15/120 to 75/60 but the one-way Mg efflux was not significantly affected. The major factor contributing to the reduced net efflux of Mg appeared to be the increase in one-way Mg influx from 15 to 4 0 /tmol/l -5 h which took place in response to the rise in transmural p.d. observed (I to mv, blood positive). The results obtained with the remaining four sheep were similar in that a change from a low to a high K/Na ratio within the rumen pouch was associated with an increased p.d. across the wall of the pouch and a significant reduction in the net efflux of Mg from the pouch. Washed rumen. This technique was also used with solutions containing 20 mmol Ca/l, 2 mmol phosphate/l, 2 5 mmol Mg/l and either 30 or 90 mmol K/I and 90 or 30 mmol Na/l, respectively. There was a significant (P < 0-05) reduction from (6) to (5),tmol/l. min in the rate of net efflux of Mg from the solution with the high K/Na ratio relative to that with the low K/Na ratio. Moreover, there was a positive correlation (r = 0.69, n = 15, P < 0 01) between the rates of net effilux of Mg and that of Na. The influence of ammonium ion concentration on the absorption of Mg from the rumen The washed rumen technique was employed to study the interaction between ammonium ion concentration and K/Na ratio on the net efflux of Mg from the rumen. All solutions used contained the composition shown in Table 1 except that the ratio of K/Na was either 30/90 or 90/30 (in mmol/l) and ammonium ions were either absent or present at 30 mmol/l. Isotonicity between solutions was maintained using mannitol and the initial ph was 7.

9 MAGNESIUM ABSORPTION IN SHEEP In seven of the eight experiments, increasing the ammonium concentration from zero to 30 mmol/l at a constant K/Na ratio of 30/90 caused a decrease in the net efflux of Mg. The same decrease was 36% of the efflux of Mg observed from the fluid without ammonium ions. When the initial ammonium ion concentration was kept at 30 mmol/l but the K/Na ratio increased to 90/30 there was a further reduction in the efflux of Mg in all five experiments conducted. The mean percentage fall was 92% of the Mg efflux found at low K/Na ratio and zero ammonium concentration. An increase in the concentration of ammonium ions also decreased the mean net efflux of Na ions by 18% and the combination of high ammonium ion concentration and high K/Na ratio reduced the mean efflux of Na ions by 57%O from the level reached in the absence of ammonium ions and at a low K/Na ratio. Regression analysis of the net efflux of Mg with the net efflux of Na showed a significant correlation (r = 0-92; n = 14; P < 0-001). Similarly, a significant correlation was obtained between the net efflux of Mg and the net efflux of Na from five rumen pouches (r = 0 59; n = 65; P < 0 001). The gradients of the two regression lines were similar (0-018 for the washed rumen and for the rumen pouch). Changing the concentration of ammonium ions alone was without effect on the potential difference across the wall of the washed rumen but increasing the K/Na ratio from 90/30 in the presence of NH4 (30 mmol/l) led to a significant increase in p.d. from to mv (P < 0-001). The addition of ammonium ions to the rumen fluid caused no significant rise in either initial or final ph of the solution. During the experiments involving rumen pouches there was no significant change in ph during a period of study, whereas during experiments with a washed rumen the ph rose from 6-8 to 7-4 during the 2 h period despite the addition of carbon dioxide to the experimental rumen fluid. H. Martens (personal communication) found that the absorption of Mg from a washed rumen containing a solution of similar composition to that used by us, was limited by ph only if the ph was below DISCUSSION Tomas & Potter (1976) found the rate of net absorption of Mg from the sheep rumen to be mmol/d when the concentration of Mg in the rumen fluid was mmol/l. Also, Grace & MacRae (1972) and Grace et al. (1974) calculated the rate of net uptake of Mg from the forestomach of sheep to lie within the range mmol/d. From Fig. 1 it can be seen that the rate of absorption of Mg from 50 ml of a solution of Mg concentration 2 5 mmol/l retained in a rumen pouch was approximately 30,tmol in 2 h. When comparing this to absorption from an entire rumen, one can approximate the rumen and its pouch to spheres of volume 5 and , respectively. The ratio of their surface areas is 22. Thus, the comparable rate of net absorption of Mg from the entire rumen can be approximated to 8 mmol/d. This estimate is likely to be low because it takes no account of the greater size of papillae in the intact rumen compared with those in the pouch. The rates of net absorption of Mg observed using the washed rumen (which has papillae of normal size) varied from 3 to 19 mmol/d at a rumen fluid concentration of 2-5 mmol/l. It can thus be seen that the estimates of net efflux of Mg obtained by the use of the pouch or the washed rumen technique are similar to those found by other investigators using less direct methods. These comparisons have been made in the absence of an excess of K ions in ruminal fluid. It can be seen from Fig. 1 that the gradient of the relationship between net efflux of Mg and intraruminal Mg concentration is maximal over the range mmol/l. This

10 586 corresponds to the range of concentration of ultrafilterable Mg found in ovine rumen fluid by Storry (1961). Thus, any factor which results in a relatively small decrease in the concentration of ionized Mg in rumen fluid may lead to a relatively large fall in Mg absorption. The curve shown in Fig. 1 is of Michaelis-Menten form and may be described by the equation: Mg efflux = Vmax-K Mgefflux Mg where Km is the Michaelis-Menten constant, [Mg] is the intraruminal concentration of Mg ions and Vmax is the maximal efflux. This saturable uptake may be viewed as a facilitated diffusion mechanism which involves a carrier. Ca ions may compete with Mg in this process. The energy-dependent step in this model may be located at the serosal surface and may be related to active Na transport (Steven & Marshall, 1970; Martens et al. 1978). The significant correlation between net effluxes of Mg and Na obtained using the washed rumen technique (Care et al. 1982) supports this hypothesis, as did the significant correlation (r = 0.59, P < 0001) shown by pairing the net effluxes of Mg and Na from sixty-five measurements in five sheep with a rumen pouch. Also, when cannulated sheep were made Na deficient, the net absorption rate of Mg from the forestomachs was significantly reduced (Kubel, 1982). It is thus possible that the mechanism is similar to that postulated by Palaty (1974) to account for the maintenance of intracellular Mg ion concentration in vascular smooth muscle. That is, a non-electrogenic exchange of one Mg ion uphill from intracellular to extracellular fluid coupled to the downhill flux of two Na ions in the opposite direction. This mechanism is dependent upon the maintenance of a functional Na pump and would thus be abolished by ouabain or by the absence of Na from the extracellular fluid. An important component of the process of net transport of Mg is the influx to the rumen contents, calculated from experiments using 28Mg. This influx of Mg was increased in the presence of a high K/Na ratio in ruminal fluid and the concomitant elevated transmural p.d.; it presumably resulted from increased diffusion from the plasma down the enhanced electrochemical gradient. The inhibitory effect of intraruminal ammonium ion on the net absorption of Mg from the rumen shown above was obtained with a concentration within the range attained by feeding young grass (30-70 mmol/l) and confirms the result obtained by Martens & Rayssiguier (1980). This effect was also associated with a reduction in Na efflux and may have been caused by a reduction in blood flow through the rumen wall (Hogan, 1962) as there was no effect on the transmural p.d. Since much of the endogenous secretion of Mg takes place in saliva it can be readily seen that if conditions within the rumen change so as to reduce the rate of net absorption of Mg, not only is dietary Mg not adequately absorbed but there is also an increased overall loss of salivary Mg. With adult animals having low stores of depletable Mg, hypomagnesaemic convulsions may then develop rapidly. REFERENCES BRAY, G. A. (1960). A simple efficient liquid scintillator for counting aqueous solutions in a liquid scintillation counter. Analytical Biochemistry 1, BROWN, R. C., CARE, A. D. & PICKARD, D. W. (1978). Magnesium absorption from the rumen of sheep. Journal of Physiology 276, 62P.

11 MAGNESIUM ABSORPTION IN SHEEP 587 CARE, A. D., FARRAR, A. R. & PICKARD, D. W. (1982). Factors affecting the absorption of magnesium from the rumen in sheep. Journal of Physiology 325, 44P. CARE, A. D. & VAN'T KLOOSTER, A. TH. (1965). In vitro transport of magnesium and other cations across the wall of the gastrointestinal tract of sheep. Journal of Physiology 177, DOBSON, A. & PHILLIPSON, A. T. (1958). The absorption of chloride ions from the reticulo-rumen sac. Journal of Physiology 140, DOBSON, A., SELLERS, A. F. & GATEWOOD, V. H. (1976). Dependence of Cr-EDTA absorption from the rumen on luminal osmotic pressure. American Journal of Physiology 231, DOWNES, A. M. & MCDONALD, I. W. (1964). The chromium-51 complex of ethylenediamine tetraacetic acid as a soluble rumen marker. British Journal of Nutrition 18, EDRISE, B. M. & SMITH, R. H. (1979). Absorption and secretion in the omasum of the young steer. Annales de Recherches Veterinaires 10, FERREIRA, H. G., HARRISON, F. A. & KEYNES, R. D. (1966). The potential and short-circuit current across the isolated rumen epithelium of the sheep. Journal of Physiology 187, FERRERIA, H. G., HARRISON, F. A., KEYNES, R. D. & NAUSS, A. H. (1966). Observations on the potential across the rumen of the sheep. Journal of Physiology 187, Foss, D. C. & BLACK, D. L. (1972). Establishment of a rumen pouch in sheep and design of a pouch perfusion system. Journal of Animal Science 34, GRACE, N. D. & MAcRAE, J. C. (1972). Influence of feeding regimen and protein supplementation on the sites of net absorption of nmagnesium in sheep. British Journal of Nutrition 27, GRACE, N. D., ULYATT, M. J. & MAcRAE, J. C. (1974). Quantitative digestion of fresh herbage by sheep. III. Journal of Agricultural Science 82, HOGAN, J. P. (1962). The absorption of ammonia through the rumen of the sheep. Australian Journal of Biological Science 14, HORN, J. P. & SMITH, R. H. (1978). Absorption of magnesium by young steer. British Journal of Nutrition 40, HYDEN, S. (1955). A turbidimetric method for the determination of higher polyethylene glycols in biological materials. Annals of the Royal Agricultural College of Sweden 22, KOMAREK, R. J., LEFFEL, E. C., BROWN, W. H. & MASON, K. R. (1959). Technique for establishment of a rumen pouch. Journal of Applied Physiology 15, KUBEL, 0. W. (1982). Thesis for the Degree of Dr Vet. Med., Tierarztlichen Hochschule, Hannover, W. Germany. MARTENS, H. (1983). Saturation kinetics of magnesium efflux across the rumen wall in heifers. British Journal of Nutrition 49, MARTENS, H., HARMEYER, J. & MICHAEL, I. (1978). Magnesium transport by isolated epithelium of sheep. Research in Veterinary Science 24, MARTENS, H. & RAYSSIGUIER, Y. (1980). Magnesium metabolism and hypomagnesaemia. In Proceedings of the 5th International Symposium on Ruminant Physiology, ed. RUCKEBUSH, Y. & THIVEND, P., pp Lancaster: M.T.P. Press. PALATY, V. (1974). Regulation of the cell magnesium in vascular smooth muscle. Journal ofphysiology 242, PFEFFER, E., THOMPSON, A. & ARMSTRONG, D. G. (1970). Studies on intestinal digestion in sheep. British Journal of Nutrition 24, PHILLIPSON, A. T. & STORRY, J. E. (1965). The absorption of calcium and magnesium from the rumen and small intestine of the sheep. Journal of Physiology 181, SCOTT, D. (1966). The effects of sodium depletion and potassium supplements upon electrical potentials in the rumen of the sheep. Quarterly Journal of Experimental Physiology 51, STEVEN, D. H. & MARSHALL, A. B. (1970). Organization of the rumen epithelium. In Physiology of digestion and metabolism in the ruminant. Proceedings of the Third International Symposium, Cambridge 1969, ed. PHILLIPSON, A. T., pp Newcastle: Oriel Press. STORRY, J. E. (1961). Studies on calcium and magnesium in the alimentary tract of sheep. Journal of Agricultural Science 57, TOMAS, F. M. & POTTER, B. J. (1976). The site of absorption of magnesium from the stomach of sheep. British Journal of Nutrition 36,