1. - Absorption from and secretion of magnesium into the gastrointestinal tract of sheep were investigated by following the distribution of ²⁸Mg in the ingesta along the tract after oral or intravenous administration of single doses of ²⁸Mg. Two 5-year-old wethers were given a diet of grass nuts providing 1.2 g. of magnesium daily for a period of 6 days. They were slaughtered 10 hours after receiving a single dose of ²⁸Mg by stomach tube (sheep A) or by intravenous injection ( sheep B) . ²⁸mg and total magnesium were determined in the liquid.and solid phases of the ingesta and in the mucosa of the various sections of the gastro-intestinal tract. selected tissues of sheep B were also analysed° The main sits of absorption of magnesium was the middle third of the small intestine. Magnesium was secreted into the lumen of both the abomasum and the first section of the small intestine and was reabsorbed from the lower segments of the small intestine. For both sheep the specific activities of the mucosa of those sections where absorption or secretion was not known to take place differed from that of the plasma; which was tentatively attributed to exchange of ²⁸Mg between the contents of the tract and the walls. After oral administration; the ionic ²⁸Mg and dietary magnesium did not distribute uniformly in the gastro-intestinal tract. The specific activity of the liquid phase was generally higher than that of the corresponding solid phaee, but the difference was not constant from section to section. After intravenous administration there was no evidence of appreciable exchange between the ²⁸Mg in the secretions with the stable exogenous magnesium in the lumen of the gut. There were marked variations in the specific activity between soft tissues, between bones and between parts of the same bone. The order of decreasing specific activity was bile, kidney, plasma, liver, spleen, skeleton muscle and bone. These results are discussed in relation to the use of ²⁸Mg to determine endogenous faecal magnesium excretion and availability of dietary magnesium by the comparative-balance and isotope-dilution methods and it was concluded that these techniques are of doubtful value. Samples of faeces were collected over 3 h. periods, for a total of 24 a., from 4 wethers that had been on the same 'dietary regime as sheep A and B for the previous 15 days. No evidence for a diurnal variation in the concentration of magnesium in faeces were obtained. 2. A study was made of the effect of dietary changes involving hay, grass nuts and spring herbage on the urinary magnesium excretion and the concentration of magnesium in the serum of adult wethers of various breeds. The samples of spring herbage were collected from fertilised leysp on one of which hypomagnezaemic tetany in cows and ewes had recently occurred. No dietary change produced any marked alteration in the serum magnesium levels and the values remained within the normal range, even when the tetany herbage was given for 18 days. i.k4 immediate fall in urinary magnesium excretion usually occurred after a dietary change from hay or grass nuts to spring herbage, even when the change led to increased magnesium intake. Lowest values, which occurred within 1 to 2 days of the dietary change, were followed by an increase although the magnesium intake remained constant. In some instances the excretion at the end of the period of observation reflected the difference in the magnesium content of the two rations. The reverse dietary change produced the opposite effect on urinary magnesium, namely, an immediate increase followed by a fall. The increase occurred when the dietary change led to a decreased intake, and maximal values were observed within 1 to 3 days of the change. The volume of urine excreted was highest 2 or 3 days after the diet had been changed from grass nuts or hay to spring herbage. It is not known whether these changes in urinary magnesium excretion are caused by changes in absorption or in endogenous faecal magnesium excretion. 3. The results of two experiments carried out at an interval of one year are reported, each experiment comprising a series of magnesium and calcium balance trials in which each of four Cheviot wethers was given 900, 1200 or 1500 go/day of the same batch of grass nuts for 15 day periods. For all sheep there was a highly significant rectilinear relationship between the amount of magnesium in the urine and the magnesium intake, but there were significant differences between the regression coefficients for the individual sheep. From these equations estimates of the percentage absorption of the magnesium in as nuts by individual sheep were obtained and ranged from 3.2 to 11.3 per cent. The relationship between faecal magnesium output and magnesium intake differed in the two experiments; in the first it was slightly curvilinear and in the second rectilinear. These relationships gave an estimate of 12 per cent for the availability of the magnesium in grass nuts to all the sheep. The estimation of the availability of dietary magnesium and endogenous faecal loss of magnesium by the methods based on the regression of urinary magnesium and of faecal magnesium on magnesium intake are discussed, and it is concluded that9for the former9 the u:?ine method is more sensitive and9 for the latter, neither method is accurate enough to give reliable estimates. No correlation between serum magnesium concentration and the amount of magnesium in the urine was detected but the mean value for the former was correlated with the animals' ability to absorb dietary magnesium. For all sheep the increase in age from 4. to 5 years had no effect on the efficiency .of absorption of dietary magnesium9 but the concentrations magnesium in the serum and urinary magnesium excretion were higher at 4 than at 5 years of age. No inter-relationships between magnesium and calcium metabolism at the dietary intakes investigated could be detected. Urinary and faecal calcium excretion were correlated with intake the former relationship was curvilinear and the latter rectilinear. From the latter an estimate of 6.4 per cent for the availability of the calcium in grass nuts was obtained.