A complete account of the changes in the blood picture in infancy requires a knowledge of the rates of formation and destruction of blood. Since the latter isnot easily measured, our approach has been to use the activity of bonemarrow as a measure of blood formation, and thento deduce whether changes in blood formation aloneare sufficient to account for observed changes in the blood picture. Part I of the study, therefore, consists ofan account of the marrow picture during the first three months of life. Part II concerns erythropoiesis, and the marrow findings are then considered together with the blood picture.

Matenal and Methods The data to be discussed are derived from a total of 102 specimens of marrow from 25 infantsaged from a few minutes to three months. A sample of venous blood was taken at the same time as the marrow. The infants were born in the Cambridge Maternity Hospital; the majority were chosen because their mothers had already borne a normal infant at the hospital and had volunteered to take part in the investigation; three subjects were mongols, but as their weight gains proved normal during the period under review, they were included in the series. Labour and delivery were normal, miniaial nitrous oxide and air anaesthesia being given to the mothers. Birth weights exceeded 6 lb. The cord was clamped as soon as conveniently possible, but usually some 10 to 15 seconds elapsed beforethis was done. Infants developing more than trivial infections were excluded. Sixteen of the 25 infants were completely breast fed. Weight gains were satis-factory and the infants generally were healthy and well cared for.