In his Paper concerning the activity responses of a diurnal and nocturnal lizard to light and temperature fluctuations, Evans (1966) points out that diurnal ^* rhythms have been found in most vertebrate classes. However, it seems clear that daily ^† rhythms are to be found in virtually all living species (Hart, 1964; Pittendrigh and Minis, 1964). This is supported by Menaker (1969) who notes that daily rhythms have been found in all major groups of organisms in which a concerted search for them has been made.

Park (1940, 1941) subdivides rhythmic phenomena into two main categories:

In the light of the great store of recent descriptive work, most workers are in agreement that daily rhythms are endogenous (Pittendrigh, 1960) and that although they are not necessarily a direct response to environmental changes, are frequently correlated with them (Cloudsley-Thompson, 1961; Hamner and Enright, 1967). Cloudsley-Thompson (1961) considers that under natural conditions, several external factors are probably active at the same time, generally with one in particular being the ruling factor of an animal's periodicity.

Brown and his co-workers (1954-6) believe that some extraneous force such as cosmic ray showers, barometric pressure, conductivity or ionisation of the atmosphere or changes in the earth's geomagnetic field may be involved as synchronisers of natural rhythmicity with the environmental day-night cycle. Work by Pittendrigh (1961), however, has led him to believe that light and temperature are the only two variables known to be coupled to the living oscillation. Although there are some indications (Aschoff, 1958) in favour of Bruce's (1960) theory that some other type of periodically repeated stimulus may cause a persistent rhythm to become synchronised with the entraining cycle, Aschoff (1963) considers that there has been no adequate demonstration of an effective zeitgeber (phase-setting factor) other than light and temperature. Of these two, he further considers, as does Cloudsley-Thompson (1961), that light is the most common and most important factor.

Evidence for the endogenicity of rhythms comes from studies which show that daily rhythms persist with periods other than those of environmental factors when organisms are placed into constant conditions (Roberts, 1960; Sollberger, 1965; Hamner and Enright, 1967; Menaker, 1969; and others). Experimental results have indicated that although perfect constant conditions are probably impossible to establish in the earth environment owing to the difficulty in excluding such variables as barometric pressure, magnetic field and ionisation of the air (Menaker, 1969), constant levels of light intensity and temperature are sufficient to demonstrate the endogenous nature of biological rhythms (Pittendrigh and Bruce, 1957; Bünning, 1958; Pittendrigh, 1960; Menaker, 1969). A major property of biological rhythms under these conditions is the deviation of period length from the exact 24-hour cycle of the natural day (Lohman, 1967), hence the term ‘circadian’, derived from the Latin (circa = about, and dies = a day) (Halberg, 1959).

Further evidence for endogenicity comes from the results of several workers (Harker, 1953; Aschoff and Meyer-Lohmann, 1954; Pittendrigh, 1954; Folk, 1955; Hoffmann, 1955) who have initiated circadian rhythmicity in laboratory-reared organisms which had never experienced environmental rhythms of 24-hour periodicity.

Marler and Hamilton (1966) and Menaker (1969) point out that individuals within a given species will exhibit small differences in the length of the natural period. Hoffmann (1957), for example, showed that lizards hatching in constant darkness and temperature had individual differences in the period of their activity rhythm. This is strong evidence against external control of the period of the rhythm since all individuals were subject to the same conditions.

Light as a Zeitgeber

The efficacy of light as an entraining agent can be demonstrated in an environment with no temperature periodicity. Under an artificial light: dark (12: 12) regime at constant temperature, Roberts (1962) found that the rhythm of the cockroach always attains a steady state whose period is 24 hours and whose phase is such that activity begins at, or shortly after, the light-to-dark transition. The primary onset of activity is closely correlated with the ‘dusk’ transition.

Concerning the natural period of the circadian oscillation, Jegla and Poulson (1968) consider that where this differs from 24 hours, factors such as dawn or dusk provide the proper phase relationship between the circadian periodicity of the species and its environment; that is, the rhythm is entrained each day as the photoperiod changes. Cloudsley-Thompson (1961) points out that synchronisation with environmental periodic changes cannot be achieved both at dawn and at dusk, as the time of each of these is altering. He considers that the synchroniser tends to be the dusk, in the case of nocturnal forms, dawn in that of diurnal forms. Bennett (1954) noted that during page 126 periods of shorter day length the greater activity of clams occurred earlier in the day than it did during the times of the year when days are longer. DeCoursey (1960) with flying squirrels, and Rowley (1957) and Holler and Marsden (1970) with rabbits recorded similar results. Such results, as with that from Hirai (1969), who found adult eclosion of Hyphantria cunea to be promoted by the change from light to dark, point to the effectiveness of dusk as a phase-setting factor. Kavanau (1962), on the other hand, found that for deer mice, which are nocturnal, dusk is sometimes ignored, the dawn changes usually being the more compelling!

Much experimental work has been done under constant light conditions. Such conditions may positively or negatively affect the amplitude of the rhythms, and may also affect the period length (Harker, 1958). Roberts (1960), for example, although finding no obvious correlation between period length and intensity of illumination, in two species of cockroaches, did note that the period was markedly lengthened in constant light as compared with constant darkness. Whereas the activity period of the white-footed mouse is also lengthened (Johnson, 1939), that of the lizard Cnemidophorus sexlineatus is shortened in constant light (Barden, 1942).

In 1960, Aschoff formulated the ‘Circadian Rule’ that in light-active animals: (1) spontaneous frequency; (2) the ratio of activity time to rest time, and; (3) total activity, all increase with increasing intensity of continuous illumination. Harker (1964) pointed out that the intensity of the light in a constant environment has a considerable effect on the length of the free-running period. Correspondingly Aschoff's hypothesis, which has been extended to include intensity effects and is now being widely called ‘Aschoff's Rule’, states, in its modified form, that an increase in the intensity of constant light causes a lengthening of the period for a nocturnal organism and a shortening of the period for a diurnal organism (Hoffmann, 1965). The rule clearly holds in those cases cited by Aschoff (1960). These include the activity rhythms of chaffinches and the lizard Lacerta sicula (Hoffmann, 1960). Similarly flying squirrels and house mice, which like deer mice are nocturnal, show an increased period length in brighter light (Aschoff, 1960; Hoffmann, 1960, 1965), as does also the mosquito, Aedes aegypti, which also shows corresponding increases in its level of activity (Taylor and Jones, 1959).

Although the validity of ‘Aschoff's Rule’ has been demonstrated in a wide variety of species (Hoffmann, 1965), rhythms of other species, which have been studied over long periods of time, show that there are exceptions (Harker, 1964). Some of the more recent studies demonstrating the latter include those of Imlay (1968) and Youthed and Moran (1969), who worked with the clam (Elliptio complanatus) and larvae of the ant-lion respectively. Marler and Hamilton (1966) state that diurnal animals such as most birds react to increased light intensities in the same way as lizards, which, page 127 considering the paucity of work done on lizards when compared with insects, birds and small mammals (Menaker, 1969), would seem a rather unfounded comparison as yet. This is especially so in the light of recent work by Cloudsley-Thompson (1967) with the Nile monitor and the author with the gecko (Hoplodactylus pacificus), in which both species have been shown to disobey ‘Aschoff's Rule’.

During his work on Peromyscus, Johnson (1939) noted that light may have an inhibiting effect on the activity of a nocturnal animal. Munn (1950) recorded the same phenomenon in rats. Sufficient evidence from other work led Harker (1958) to write that light partially or completely inhibits movement and other activities in some animals. She noted also that although arthropods appear to be the only group from which the following effect is recorded, continuous darkness may also be inhibitory.

An important comment from Marler and Hamilton (1966) concerns the variable effect that lighting conditions may have on individual animals. Hoffmann (1957), for example, found a strong individual variation in period length in adult lizards (Lacerta sicula, L. agilis, L. viridis) measured under constant conditions.

Temperature as a Zeitgeber

Regarding the importance of temperature as a factor in daily activity Bünning (1931), Wahl (1932) and Kalmus (1934) found the period length to be remarkably independent of temperature under steady conditions. Later work has confirmed these early findings in a numbr of widely varying species, for example: tortoises (Cloudsley-Thompson, 1970); Perognathus intermedius (Stewart and Reeder, 1968); Bufo fowleri and B. americanus (Higginbotham, 1939); Uca (Brown and Webb, 1948); and Carcinus (Naylor, 1960).

On the other hand, studies of lizards (Hoffman, 1957) and many other species (Sweeney and Hastings, 1960) have consistently revealed small but distinct shifts in period length within certain ranges of temperature change. Bustard (1968) has found temperature above 25-26° C. delays initiation of evening activity of the nocturnal gecko, Diplodactylus vittatus, whereas a fall in evening temperature to below 13-17° C. greatly curtails evening activity.

Marler and Hamilton (1966) consider the relative temperature independence of circadian rhythms to be significant under natural conditions. They point out that if circadian rhythms serve primarily to concentrate appropriate behaviour at certain times of day, change with temperature would hinder accurate timing. On the other hand, as the rate of metabolic processes is so closely linked with temperature, it is surprising that rhythms are not also affected by temperature (Harker, 1958; Marler and Hamilton, 1966).

Although constant temperatures cause little change in period length, it has been found that a regular cycle of temperature change page 128 is quite effective in synchronising some circadian rhythms (Roberts, 1960; Aschoff, 1963). Whereas it had previously been believed that regular temperature fluctuations had an effect on the cockroach rhythm (Cloudsley-Thompson, 1953; Harker, 1956), Roberts (1962) has shown this to be false. He considered the efficacy of temperature as an entraining agent to be noteworthy on ecological grounds: (1) the phase of the entrained rhythm is approximately coincident with the high point of a temperature cycle and with dusk, in a light-dark regime; and (2) these two distinct entraining agents operate simultaneously in nature and give non-conflicting information, sunset and the high point of the temperature curve being roughly coincident.

Bentley, Gunn and Ewer (1941) showed that the activity rhythm of the spider beetle Ptinus tectus, though gradually lost in continuous illumination, could be reinstated by periodic exposure to high (23° C.) and low (17° C.) temperatures. Other species which become synchronised with temperature fluctuations include the lizards, Sceloporus magister (Taylor and Tschirgi, 1960) and Uta stansburiana (Evans, 1966). Hirai (1969) has found that a drop in temperature induces earlier eclosion in Hyphantria.

While temperature entrainment has, therefore, been well substantiated for such poikilotherms as insects and lizards, it has not been demonstrated conclusively for mammals (Stewart and Reeder, 1968). DeCoursey (1960) with flying squirrels, and Bruce (1960) with hamsters, were unable to find any evidence for temperature entrainment. One should note, however, that Browman (1943) and Calhoun (1944) have shown a temperature cycle to determine phase-setting in the rat.

Cloudsley-Thompson (1961) pointed out that in all organisms investigated up till then, there appeared to be a critical temperature at which rhythms cease. Furthermore, he contended that although the period of a rhythm may be relatively unaffected by temperature, its amplitude will show a normal physiological temperature dependence. Mark and Kayser (1949) found this to be the case in Lacerta agilis and L. muralis, as did the author with Hoplodactylus pacificus. Earlier, Higginbotham (1939), working with toads, had discovered that with an increase of 10° C., a doubling or tripling of the amount of activity occurred.

Harker (1958) showed that sudden large changes or very low temperatures can alter the period of a daily rhythm which is not temperature sensitive within a normal temperature range. Phase shifts are reported to occur in Uca (Stephens, 1957), and Periplaneta (Bünning, 1958), when the temperature is lowered to the 0-10° C. range for an interval of 12 hours or less.

Temperature and light may also have interacting effects (Marler and Hamilton, 1966). For example, Enright (1966) found that the extent to which the free-running activity of the house finch is retarded by low temperatures, depends on the intensity of the constant light.