Rats mated under normal lighting conditions of a 24 h day were transferred to a sound‐proof room with period lengths (T) of 24 h (12 h light and 12 h dark, 12L:12D), or 22 h (11L:11D). During the first 15 d of pregnancy the rats were kept together in a large enclosure and the locomotor activity of the group was recorded. From day 15 they were kept separately, in cages, and from day 21 the number of litters born was noted every 3 h.
On 12L:12D, rats displayed well defined periods of activity which were closely related to photoperiod. Mean activity was greatest in the dark, and overall activity declined after day 13 of gestation. On 11L:11D, coupling of locomotor activity to the dark phase was retained initially, but was not sustained throughout the period of study. Activity in the light was comparable to that in the dark at the time when the light:dark cycles of 11L:11D became reversed relative to those of 12L:12D.
Rats on a 24 h day had a peak time of delivery at 519 h after mating (21·6 cycles of 12L:12D) and those on a 22 h day at 527 h (24·0 cycles of 11L:11D; difference, P 〈 0·002). The 11L:11D cycle therefore increased rather than decreased the duration of pregnancy, implying that elapsed time rather than the number of light:dark cycles experienced has the greater influence on the length of gestation. Of the animals on a 24 h day 76% had their litters in the light phase of the cycle, whereas only 36% of those on a 22 h day gave birth in the light (P 〈 0·02).