Relating back to research on sleep of French train drivers carried out by Foret and Lantin (1972), work schedules from a random sample of 180 successive on- and off-duty days of 4 train drivers were analysed. Work blocks cover 4 to 8 consecutive days, with off-duty times spent at home or away from home. Within these work blocks it appears that in 84 per cent of the cases, individual starting and finishing times of service fall earlier from one service to the following. On the whole they tend to sweep continuously the nycthemere counterclockwise, which results into less than 24 hours almost-periodicity of the work rest cycles. In 95 per cent of the work blocks, estimated periods are less than 24 hours, ranging in these cases from 19 hours to 23.60 hours (sample mean: 20.68 hours). Mean service duration observed in the sample is 7 hours. Mean effective driving task corresponds to 60 per cent of that time.
The general hypothesis is made that such an artificial application of a work rest cycle which is less than 24 hours provides the least propitious conditions for development of so-called adaptive responses, at least as far as duration of sleep and intrasleep patterns are concerned, if subjects are at the same time submitted to circadian influences of environmental and socialZeitgeber.
Key words Work schedules - REM sleep - Zeitgeber - Biorhythms - Train drivers