This paper develops several propositions concerning the lability of the amplitude of Drosophila circadian pacemakers. The first is that the amplitude of the pacemaker's motion, unlike its period, is markedly temperature-dependent. The second is that latitudinal variation in pacemaker amplitude (higher in the north) is responsible for two very different sets of observations on Drosophila circadian systems at successively higher latitudes. One of these is a cline in D. auraria's phase-shifting response to light, which steadily weakens in a succession of more northerly strains. The other, concerning D. littoralis in the very far north, is a cline in the rate at which eclosion activity becomes arrhythmic (the circadian rhythm damps out) in constant darkness; damping is faster in the north. The third proposition concerns a plausible selection pressure for the cline in pacemaker amplitude that we propose underlies the two directly observed clines. Two points are emphasized: (1) The amplitude of the pacemaker's daily oscillation declines as the duration of the entraining light pulse (photoperiod) is increased; and (2) the duration of the daily photoperiods throughout the breeding season is steadily increased as one moves toward the poles. Selection for conservation of pacemaker amplitude (during the breeding season) would produce the latitudinal cline we propose. The fourth, and final proposition is that since the amplitude of the pacemaker's daily motion responds systematically to change in photoperiod, amplitude is clearly one way--and a temperature-dependent way--in which insect circadian systems may sense seasonal change. These propositions concerning the temperature and latitude dependence of pacemaker amplitude may be relevant to a wider array of circadian pacemakers than Drosophila.