The first investigators to succeed in producing magnesium (Mg) deficiency experimentally were Orent and his coworkers,' who reported in 1934 that the rats used for their study showed a depletion of bone magnesium. Subsequent investigators2 have reported that the soft-tissue content of magnesium is not appreciably altered even in severe deficiency. When zsMg became available several years ago, we made a preliminary study of the dynamics of magnesium transfer in the tissues of rabbits maintained on a Mg-deficient diet for 30 days.3 The results of this study are reviewed in the present report, which also describes subsequent studies made in rabbits maintained on a diet extremely low in magnesium for four to six weeks. In addition to studying changes in blood constituents in such animals, we have done some preliminary in vitro studies on the intestinal transfer of magnesium in Mg-deficient and normal rabbits.
THE EXTERNAL BALANCE AND Mg, STUDY
Material and MethodsAdult male domestic rabbits were placed in individual stainless steel metabolism cages and were given unrestricted quantities of tap water containing 0.6 milliequivalent (mEq) of magnesium per liter. The basal diet used for rats by Mackenzie and Mackenzie4 was supplemented by the salt mixture of Hubbell and his coworkers,5 except that the magnesium salts were omitted. The composition of this diet, which contains 6.6 mEq of magnesium per kilogram, has been described previously.The z8 Mg was received as MgCI2 in concentrated HCI; the specific activity was on the order of 8-10 pc/mEq. The concentration of magnesium in the solution injected was 0.4 mEq/ml, and each injection contained 2 mEq.The procedures for collecting the samples for the external balance study, for measuring the exchangeable magnesium content ( Mge), and for assaying radioactivity have been described previ~usly.~ The magnesium concentration in urine and serum6 and in tissues7 was determined by a modification of the molybdivanadate method for phosphate.In the statistical analyses, the t test* was used to determine the significance either of the mean differences or of the differences between group means. A P of <0.01 was considered significant.This experiment was conducted as follows. During a control period of three days the eight rabbits in the test group were fed a balanced stock diet of compressed pellets, containing 172 mEq of magnesium per kilogram. Base-line data for body weight. serum magnesium concentration, and urinary excretion of 144