Summary. The secretion of prolactin comes under important seasonal variations in sheep. Thus, in the ram the highest levels are in summer and the lowest in winter. These variations of secretion are under light control because illumination of long duration (long photoperiods) produces high prolactin levels (220 -L 2 ng) ; short photoperiods produce low levels (12 ! 2 ng).Using the hypothesis of Bunning, we study a photoinducible phase during diurnal rhythm. For this purpose, rams preconditioned either in short photoperiods or in long photoperiods, receive 8 hours of illumination in two fractions of 7 and 1-hour durations. The beginning of the 7-hour period is considered to be a subjective dawn, and the 1-hour light pulse is given at various times after dawn : at hours 11, 14, 17 and 20.The group receiving the light pulse during hour 17 after dawn has significantly higher prolactin levels than all the other groups. For example, in rams preconditioned in short photoperiods and after 5 weeks of light pulse treatment, the mean prolactin levels are 71 ! 4 ng (hour 11)) ; 68 ! 5 ng (hour 14) ; 225 ! 22 ng (hour 17) and 110 ! 16 ng (hour 20). The (Pelletier, 1973) (Tucker et al., 1973 ;Schams and Reinhardt, 1974) and in goats (Buttle, 1974 ;Hart, 1975). One of us (Ravault, 1976) (Kann, 1971) (Hamner, 1963(Hamner, , 1964Farner, 1965 ;Menaker, 1965 ;Follett and Sharp, 1969 ;Follett, 1973 ;Lofts and Lam, 1973 Ronnekleiv, Krulich and MacCann, 1973 ;Simionescu, Sahleanu and Oprescu, 1974 ;Kizer et al., 1975), in the pregnant rat (Butcher, Fugo and Collins, 1972) or in the pseudopregnant rat (Freeman and Neill, 1972), in the ewe during the luteal phase (Davis and Borger, 1972), in the lactating cow Convey, 1970, 1972), in the rhesus monkey (Quadri and Spies, 1976) and in man (Sassin et al., 1972 ;Vanhaelst et al., 1973 ; Rubin et at., 1975 (Buttle, 1974).We have studied the evolution of plasmatic prolactin level during the diurnal cycle in rams before the beginning of light pulse treatment, using either long photoperiods (June 8) or short photoperiods (December 15), and then on the same rams about five weeks after the beginning of the light pulse treatments.In all groups, the mean prolactin level was higher in the dark than in the light phase. For example, in the pretreatment period ( fig. 6), the mean prolactin level in 30 rams was 190 ! 9 ng/ml (long photoperiods) and 32 ! 4 ng/ml (short photoperiods) during the dark, compared with 112 ! 5 ng/ml and 6 ± 1 ng/ml during the light phases. The results after about one month of light treatment are shown in table 2.This difference is due to the sharp increase in prolactin in the hour after the change from light to dark ( fig. 7, 8). A high level is subsequently maintained for 2-3 hours, then falls gradually. The short duration of this peak explains why several authors have not found diurnal variations : blood sampling was not frequent enough.Other low but significant peaks of prolactin were observed either at the end of the dark phase on rams preconditioned to short p...
