“…In most cases the results are based upon analysis of records of premature infants, or of stillbirths or deaths, without any control groups, nor is there usually any differentiation by booked and emergency admissions, primiparity and multiparity, or any clear indication of the composition of the hospital sample in relation to the general population. Anderson, Brown and Lyon (1941) in a controlled survey of a year's births in Cincinnati General Hospital, give toxaemia, uterine bleeding and syphilis as the most important complications leading to prematurity. Diddle and Plass (1942) also specify toxaemias and syphilis as being of major importance.…”