CASES of hypocalcaemia may be divided into three groups. Firstly, there are the various types of low-calcium tetany. Secondly, there are cases of hypocalcaemia associated with low serum-protein, in which no symptoms of tetany occur; this condition may be present in nephrosis, chronic nephritis with oedema, kala-azar and some other conditions. Thirdly, there are cases of chronic nephritis with uraemia, in which the inorganic phosphorus of the serum is increased, and in which the serum-proteins may be normal or only moderately reduced; in some instances neuromuscular irritability and carpopedal spasm occur in addition to uraemic manifestations. One would expect that in tetany a low [Ca++] would be the significant change. The ionic calcium may or may not account for the whole of the diffusible calcium of serum, but it must be included in the diffusible fraction. It would also be expected that when the serum-calcium falls in response to a rise in the inorganic phosphate, the diffusible fraction would be primarily affected, though the change in the diffusible fraction would lead to a secondary change in the protein-bound fraction as a result of the disturbance of equilibrium. On the other hand, where the serum-calcium falls as a result of a fall in the serumprotein, one would expect the protein-bound fraction to be reduced and the diffusible fraction normal, and the absence of tetany would thus be explained. Now those authors who have used the concentration of calcium in the cerebrospinal fluid as a measure of the diffusible calcium of the serum have arrived at results which are inconsistent with the above views, for it has been shown that when the serum-calcium falls in tetany, the calcium of the cerebrospinal fluid is relatively little changed. Authors who have determined the diffusible calcium of the serum by ultrafiltration or dialysis have arrived at variable results, but, on the whole, work done on ultrafiltration of human sera has shown that the diffusible calcium is decreased in tetany and normal in cases of hypocalcaemia with low serum-protein. The view that the calcium of cerebrospinal fluid is equal to the diffusible calcium of serum has been attacked, but as yet only a few comparisons between the cerebrospinal fluid and the serum ultrafiltrate have been made in cases where the total serum-calcium is abnormal. At the outset of the present work the intention was to make comparisons between the calcium of serum and of cerebrospinal fluid in conditions where the