We all know that the normal human daily cycle of activity is some 7—8 hours' sleep alternating with some 16—17 hours' wakefulness and that, broadly speaking, the sleep normally coincides with the hours of darkness. Our present concern is how easily and to what extent this cycle can be modified. The question is no more academic one. The ease, for example, with which people can change from working in the day to working at night, is a question of growing importance in industries where automation calls for round-the-clock working of machines. It normally takes from five days to one week for a person to adapt to a reversed routine of sleep and wakefulness, sleeping during the day and working at night. Unfortunately, it is often the case in industries where shifts are changed every week a person may work from 12 midnight to 8 a.m. one week, 8 a.m. to 4 p.m. the next, and 4 p.m. to 12 midnight the third and soon. This means that no sooner has he got used to one routine than he has to change to another, so that much of his time is spent neither working nor sleeping very efficiently. The only real solution appears to be to hand over the night shift to a number of permanent night workers. An interesting study of the domestic life and health of night-shift workers was carried out by Brown in 1957. She found a high incidence of disturbed sleep and other disorders among those on alternating day and night shifts, but no abnormal occurrence of these phenomena among those on permanent night work. This latter system then appears to be the best long-term policy, but meanwhile something may be done to relieve the strains of alternate day and night work by selecting those people who can adapt most quickly to the changes of routine. One way of knowing when a person has adapted is by measuring his body temperature. People engaged in normal daytime work will have a high temperature during the hours of wakefulness and a low one at night when they change to night work, the pattern will only gradually go back to match the new routine and the speed with which it does so parallels, broadly speaking, the adaptation of the body as a whole, particularly in terms of performance. Therefore, by taking body temperature at intervals of two hours throughout the period of wakefulness it can be seen how quickly a person can adapt to a reversed routine, and this could be used as a basis for selection. So far, however, such a form. of selection does not seem to have been applied in practice. Why is the question of 'how easily people can get used to working at night' not a mere academic question?