“…This is contrary to what might be expected on the basis of previous reports that, if treatment was maintained for longer periods, the patient tended to take correspondingly longer to relapse (Himsworth, Morgans, and Trotter, 1947;Himsworth, 1948;Dunlop and Rolland, 1950). Douglas and Kennie (1952) have shown that the proportion of patients who relapse is much greater among those who have had treatments of less than one year before withdrawal than among patients who have been treated for one to two and a half years or more. However, inspection of their figures also suggests that, of the patients who do relapse, those who have had treatments of one year or longer tend to relapse sooner-in half the time, on an average-than those who have been treated for periods of less than one year.…”