Steers with rumen and simple duodenal cannulas were allowed to graze pasture or were given diets of dried grass or flaked maize with or without hay. For an experiment a solution or suspension of magnesium chloride, polyethylene glycol (molecular weight 4000) and either 144Ce (as cerous chloride) or chromic oxide was added to the rumen with a morning feed. Conditions in the rumen were sometimes modified by adding sodium chloride or hydrochloric acid. 2. Changes in magnesium:marker in samples of strained rumen contents with time interval after adding the dose were due partly to changes in Mg distribution between different phases. Results indicated, but not unequivocably, that negligible amounts of Mg were absorbed in the first few hours. 3. Relative recoveries of Mg and markers at the duodenum indicated that proportions of Mg intake absorbed (net) varied from approximately zero for pasture to 0.2--0.5 for flaked maize. Significant correlations between absorption efficiency and sodium:potassium in rumen contents (positive) and rumen pH (negative) were observed. 4. Steers with simple duodenal and re-entrant ileal cannulas were given a diet of flaked maize and hay supplemented with different amounts of magnesium oxide. Little net change in Mg relative to an unabsorbed marker was found between these sites even for a diet containing an Mg supplement of 8 g/kg dry matter.
