1. The locomotor activity rhythm of Periplaneta americana in alternating light and darkness is described as consisting of six stages.

  2. The effect on the suboesophageal ganglion neurosecretory cycle of a change from light to darkness at each stage of the locomotor rhythm is described, and three stages in the neurosecretory cycle are recognized.

  3. The effect on an established locomotor activity rhythm of a change to darkness at various times of day is described in terms of the immediate reaction of the animal and of the subsequent phase relations of the rhythm.

  4. The phases of the activity rhythm are not reset if the environmental change occurs during the active period. The final positioning of the phases, when the onset of darkness occurs during the non-secretory phase of the neurosecretory cycle, is dependent upon the subsequent light conditions; transient activity peaks may appear before the stable position is reached.

  5. The dependence of the neurosecretory cells on some other centre for the provision of some secretory substance, or precursor, is discussed.

It has been clearly established that the phases of the diurnal rhythms of most animals bear a specific relationship to the time of onset of darkness, or light, when these environmental factors are alternated within 24 hr. cycles. The completion of the light : darkness cycle is not, however, always necessary for phase determination. One short period of light will serve to establish a rhythm in some arrhythmic animals (Brett, 1955; Harker, 1958); the environmental change in this case contains no information concerning period, and the establishment of a 24 hr. response is strong evidence for an endogenous control of periodicity.

If a clear rhythm is already being shown by an animal the effect of a single alteration in the time of exposure to light or darkness might be expected to give information about the properties of the controlling mechanism of rhythmicity. Pittendrigh & Bruce (1957) have discussed the significance of the effect of single light perturbations on the resetting of phase in the eclosion rhythm of Drosophila, and have postulated control by an innate mechanism consisting of two coupled oscillators (Pittendrigh, Bruce & Kaus, 1958). The response of an oscillator to a non-periodic disturbance may give rise to transients, and it is this term which has been applied to the non-periodic peaks of eclosion which follow a single perturbation of the environment prior to the reappearance of a steady rhythm.

Some difficulties arise in the interpretation of Pittendrigh’s results because the process measured is one which takes place only once in the life-cycle of each animal. In contrast to these results transients do not always appear in the rhythm of a continuous process after a non-periodic disturbance in the environmental cycle ; phase shift may be immediate as is found in the rhythm of luminescence of Gonyaulax, a unicellular organism (Hastings & Sweeney, 1958).

A continuous-process rhythm in a more specialized animal than Gonyaulax might be supposed to involve a number of rhythms or cycles, some or all of which may be affected to a different degree by the environmental perturbation : the total effect, as measured by the reaction of the organism as a whole, might be that of a transient response. In this paper the effect of environmental perturbations on the locomotor activity rhythm of Periplaneta americana is described. This animal has been chosen because it is known that its activity rhythm is dependent on the rhythmic secretion of a hormone from the neurosecretory cells of the suboesophageal ganglion. The phases of the secretory rhythm of these cells can be measured by implanting the ganglion into an arrhythmic animal which then becomes active at the time of secretion (Harker, 1956). By using this technique the effect of an environmental perturbation on two rhythms, the neurosecretory rhythm, and the locomotor activity rhythm, can be measured.

The locomotor activity of Periplaneta over 24 hr., when light and darkness are alternated in 12 hr. cycles, follows a fairly constant form in which six stages can be recognized (Fig. 1). Stage A. Some time before the onset of darkness the level of activity may rise. Stage B. After the onset of darkness the increase in activity is sudden and marked. Stage C. A period of 2–3 hr. in which activity is high. Stage D. The level of activity decreases over a period of about 2 hr. Stage E. The activity remains at a fairly low level. Stage F. About 5 hr. after the onset of light activity is at a minimum, or may cease completely, for an hour or two : this stage appears to be directly related to the change from darkness to light, for it does not appear in the rhythm of an animal kept in continuous light.

Fig. 1.

The six stages of the locomotor activity rhythm of Periplaneta americana in alternating light and darkness. Hatched area represents dark period.

Fig. 1.

The six stages of the locomotor activity rhythm of Periplaneta americana in alternating light and darkness. Hatched area represents dark period.

In each of the following experiments the phases of the activity rhythms of the cockroaches in each group were set by keeping the animals in 12 hr. light : 12 hr. darkness for at least 3 weeks. Hereafter these conditions will be termed ’normal’, and a rhythm following the stages described above a ’normal’ rhythm. The activity of two animals at a time was recorded in each of the new conditions, and each experiment was repeated five times. As a control the activity of one group of animals was recorded in the normal conditions at the same time as each of the experiments. Activity was measured using a photo-transistor recorder working on very dim red light (Brown, 1959), except in the case of two animals in each experiment whose activity was recorded by a direct method described in a previous paper (Harker, 1956). The two methods gave similar results in all experiments.

Group A

From previous experiments (Harker, 1958) it is known that the secretory activity of the suboesophageal ganglion starts at the beginning of the active phase of the locomotory rhythm, and this in turn corresponds with the time of onset of darkness. In order to measure the effect of a change from light to darkness on the secretory cycle during the various stages of the activity rhythm groups of cockroaches were placed in darkness every 2 hr. over a 24 hr. period. Thirty minutes after the onset of darkness the suboesophageal ganglia were dissected out and implanted into arrhythmic headless animals, whose subsequent activity was recorded.

The times at which the arrhythmic animals became active, a measure of the time of neurosecretion, are shown in Fig. 2.

Fig. 2.

The time of secretion from the suboesophageal ganglion neurosecretory cells after onset of darkness at times out of phase with the activity rhythm. The arrowed line represents the time at which the dark period began on the previous day

Fig. 2.

The time of secretion from the suboesophageal ganglion neurosecretory cells after onset of darkness at times out of phase with the activity rhythm. The arrowed line represents the time at which the dark period began on the previous day

From the results (also confirmed by those in the following section), it appears that there is a stage in the neurosecretory cycle when secretion will take place whatever the external conditions, a stage when secretion will take place if there is a change from light to darkness (‘possible’ secretion), and a stage when no secretion will take place even after such a change (‘impossible’ secretion). The stages of the cycle are represented diagrammatically in Fig. 3.

Fig. 3.

Diagrammatic representation of the neurosecretory cycle in relation to time of darkness and locomotor activity.

Fig. 3.

Diagrammatic representation of the neurosecretory cycle in relation to time of darkness and locomotor activity.

If a change from light to darkness occurs at a time when secretion is already taking place the phases of the cycle appear not to be reset in any way : secretion next occurs at the ’normal’ time. If the onset of darkness occurs at a time when secretion would not normally be taking place, but when the cells are able to respond to the stimulus, then the cycle is reset, so that secretion next occurs 24 hr. after the stimulus. If secretion cannot take place at the time of the stimulus then the neurosecretory cycle appears to be unaffected, and secretion continues to take place at the ‘normal’ time.

Group B

Further experiments were carried out under a variety of conditions as detailed below. The results are shown in Figs. 4–7, in which the lines b—e refer to the conditions so labelled hereunder. Line a in each figure is the result of a control experiment in which the ‘normal’ conditions, those in which all animals had previously been maintained, were continued.

Fig. 4.

The activity of cockroaches when: (a) onset of darkness occurs at the ‘normal’ time—control experiment ; (b) darkness occurs 2 hr. eartier than previously experienced, new tight : dark phases continued ; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. dark period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous tight. Ø, Time at which ganglion removed and implanted into arrhythmic animal. ○−○, The activity of the implanted animal.

Fig. 4.

The activity of cockroaches when: (a) onset of darkness occurs at the ‘normal’ time—control experiment ; (b) darkness occurs 2 hr. eartier than previously experienced, new tight : dark phases continued ; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. dark period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous tight. Ø, Time at which ganglion removed and implanted into arrhythmic animal. ○−○, The activity of the implanted animal.

  1. Onset of darkness 2 hr. earlier than normal (Fig. 4).

    • Control.

    • Normal 12 hr. darkness : 12 hr. light cycle set forward by 2 hr.

    • Onset of darkness 2 hr. earlier than normal followed by continuous darkness.

    • Onset of darkness 2 hr. earlier than normal, initiating a cycle of 4 hr. darkness: 20 hr. fight.

    • Onset of darkness 2 hr. earlier than normal followed by continuous light.

      As an additional experiment, under some conditions, the suboesophageal ganglion was removed 2–4 hr. after the onset of darkness and implanted into an arrhythmic animal whose subsequent activity was recorded in continuous light (Figs. 4 c, 5b, 6c, 7b).

  2. Onset of darkness 7 hr. earlier than normal (Fig. 5). a, b, c, d, e as under 1.

  3. Onset of darkness 4 hr. later than normal (Fig. 6). a, b, c, d, e, as under 1.

  4. Onset of darkness 8 hr. later than normal (Fig. 7). a, b, c, d, e, as under 1.

Fig. 5.

The activity of cockroaches when : (a) onset of darkness occurs at the ‘normal’ time—control experiment ; (b) darkness occurs 7 hr. earlier than normal, new light : dark phases continued ; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. dark period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous light.

Fig. 5.

The activity of cockroaches when : (a) onset of darkness occurs at the ‘normal’ time—control experiment ; (b) darkness occurs 7 hr. earlier than normal, new light : dark phases continued ; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. dark period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous light.

Fig. 6.

The activity of cockroaches when: (a) onset of darkness occurs at the ‘normal’ time—control experiment; (b) darkness occurs 4 hr. later than normal, new light:dark phases continued; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous light.

Fig. 6.

The activity of cockroaches when: (a) onset of darkness occurs at the ‘normal’ time—control experiment; (b) darkness occurs 4 hr. later than normal, new light:dark phases continued; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous light.

Fig. 7.

The activity of cockroaches when: (a) onset of darkness occurs at ‘normal’ time—control experiment ; (b) darkness occurs 8 hr. later than normal, new light : dark phases continued ; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous light.

Fig. 7.

The activity of cockroaches when: (a) onset of darkness occurs at ‘normal’ time—control experiment ; (b) darkness occurs 8 hr. later than normal, new light : dark phases continued ; (c) new time of darkness followed by continuous darkness ; (d) 4 hr. period beginning at new time ; (e) 4 hr. dark period at new time followed by continuous light.

The following points arising from the results seem worthy of special mention.

(i) When the animals are kept in continuous darkness after the environmental perturbation the beginning of activity on successive days is 24 hr. later than the beginning of the previous active period, whether the previous active period resulted from the new or the old environmental condition. This is noticeable, for example, in the experiment in which the onset of darkness came 4 hr. later than normal (Fig. 6 c) ; activity continued after this stimulus for a longer period than normal, but the timing of the subsequent peaks is related to the beginning of the normal peak, and not to that of the second peak produced by the new stimulus.

This result is in keeping with the finding that the secretory cycle is not reset by a stimulus occurring during the secretory phase (Fig. 2 f).

(ii) When the onset of darkness came 8 hr. later than usual (Fig. 7,b), or 7 hr. earlier than normal (Fig. 5 b), the animal did not respond immediately by becoming active but the subsequent activity peak occurred 4–6 hr. earlier than normal.

In the experiment of Fig. 76 the animals implanted with ganglia removed 4 hr. after the onset of darkness also showed an activity peak 4–6 hr. earlier than normal. This shows that although neurosecretion is not immediately evoked (the onset of darkness falling within the stage of ‘impossible’ secretion) there is yet some effect which brings forward the next episode of neurosecretion by 4–6 hr. In the comparable experiment of Fig. 2g in which the ganglia were removed 30 min. after the onset of darkness, the next episode of neurosecretion was not brought forward. Nor was it brought forward in the experiment of Fig. 5 A, in which the ganglia were removed 2 hr. after the onset of darkness. It appears that the effect of the onset of darkness in bringing forward the next episode of neurosecretion is a long-term one, in that it requires that the ganglia remain in situ for some hours.

If the experiment of Fig. 7b (unoperated animals) were continued for several days and the episodes of neurosecretion were successively brought forward, this would be equivalent to a shortening of the period of the neurosecretory cycle. The result could be that eventually the onset of darkness would fall within the period of ’possible’ secretion. When this occurred there would be an abrupt shift of the activity peak so as to coincide with the onset of darkness. This in fact is observed to happen (Fig. 8).

Fig. 8.

The peaks of activity of cockroaches when (a) onset of darkness occurs at the ‘normal ’ time— control experiment ; (b) darkness occurs 8 hr. later than normal, new dark : light phases continued for several days.

Fig. 8.

The peaks of activity of cockroaches when (a) onset of darkness occurs at the ‘normal ’ time— control experiment ; (b) darkness occurs 8 hr. later than normal, new dark : light phases continued for several days.

The effects described above are similar to the transient effects discussed in the Introduction ; as was there suggested they appear to have arisen through the interaction of two factors.

(iii) The length of the active period is nearly constant whether 4 or 8 hr. darkness is experienced, except in those cases in which the peak of activity coincides with the time of the inhibitory effect of a change from darkness to light (Figs, 4d, 5d).

The length of the active period is also the same even when the animal has recently been active, as, for instance, when the dark period begins 4 hr. later than normal (Fig. 6b).

These results suggest that either the form of the neurosecretory cycle is not dependent upon the exhaustion of some substance at the end of the secretory period, or that the replacement of the secretory material (or of some precursor) is very rapid.

There is some evidence in support of each of these alternatives. The persistence for some days of a 24 hr. secretory cycle in isolated ganglia points to a certain independence of the secretory cells. On the other hand, after a change in the environmental cycle the timing of the next episode of neurosecretion depends upon the length of time for which the suboesophageal ganglion has been left with its normal connexions (§ii) ; this suggests that some other centre may be involved in supplying a substance concerned with secretion, or that the secretory cells are not entirely independent. A study of this problem is in progress.

This work was supported by a special research grant from the Department of Scientific and Industrial Research, which is gratefully acknowledged. I wish to thank Prof. V. B. Wigglesworth for his encouragement.

Brett
,
W. J.
(
1955
).
Persistent diurnal rhythmiaty in Drosophila emergence
.
Ann. Ent. Soc. Amer
.
48
,
119
.
Brown
,
R. H. J.
(
1959
).
A simple activity recorder
.
J. Inst. Physiol
.
3
,
125
.
Harker
,
J. E.
(
1956
).
Factors controlling the diurnal rhythm of activity of Periplaneta americana L
.
J. Exp. Biol
.
33
,
224
.
Harker
,
J. E.
(
1958
).
Diurnal rhythms in the animal kingdom
.
Biol. Rev
.
33
,
1
.
Hastings
,
J. W.
&
Sweeney
,
B. M.
(
1958
).
A persistent diurnal rhythm of luminescence in Gonyaulax polyedra
.
Biol. Bull., Woods Hole
,
115
,
440
.
Pittendrigh
,
C. S.
&
Bruce
,
V. G.
(
1957
).
An oscillator model for biological clocks
.
In Rhythmic and Synthetic Processes in Growth
, pp.
75
.
Princeton University Press
.
Pittendrigh
,
C. S.
,
Bruce
,
V. G.
&
Kaus
,
P.
(
1958
).
On the significance of transients in daily rhythms
.
Proc. Nat. Acad. Sci., Wash
.,
44
,
965
.