Since a review of the conditions under which siderocytes appear will support a hypothesis that it is an ageing erythrocyte at least as strongly as Gruneberg's $(1941a)$ theory that it should be considered as a young cell, a search was made in blood films of stored mammalian blood and large numbers of siderocytes were found. The conditions affecting the rate of appearance of the siderocytes were studied, and it was found that adverse conditions would hasten their appearance. The relationship of the siderotic material to the 'easily split' blood iron was also considered, and it seems probable that both are derived from a special and identical fraction of 'haemoglobin', and that this phenomenon is related to an intracorpuscular bile pigment formation. The occurrence of siderocytosis after the ingestion of acetyl phenylhydrazine by a 'normal' human being was followed, and a close correlation between siderocytosis, erythrocyte destruction and urinary siderosis as described by Peyton Rous (1918) is shown. The siderocyte extrudes its siderotic granules and reverts to a state at present morphologically indistinguishable from the normal erythrocyte, but appears to be susceptible of phagocytosis at this stage. The application of siderocyte counts to clinical medicine is suggested.