SUMMARY1. Thirty-six subjects in an isolation unit were subjected to time shifts of 12 hr, or of 8 hr in either direction.2. The rhythms of body temperature and excretion of eight urinary constituents were studied before and after the shift, both on a usual nychthemeral routine and during 24 hr when they remained under constant conditions, awake, engaged in light, mainly sedentary activity, and consuming identical food and fluid every hour.3. The rhythms on nychthemeral routine were defined by fitting cosine curves. On constant routine the rhythm after the shift was cross-correlated with the original rhythm, either with variable delay (or advance) or with an additive mixture between this variably shifted rhythm and the unshifted or a fully shifted rhythm. The process yielding the highest correlation coefficient was accepted as the best descriptor of the nature of adaptation.4. A combination of two rhythms was observed more often for urinary sodium, chloride and phosphate than for other variables.5. Adaptation appeared to have proceeded further after westward than eastward shifts, and this difference was particularly noticeable for urinary potassium, sodium and chloride.6. Partial adaptation usually involved a phase delay, even after an eastward shift when a cumulative dAlay of 16 hr would be needed to achieve full adaptation and re-entrainment. 7. Observations under nychthemeral conditions often gave a false idea of the degree of adaptation. In particular, after an eastward shift the phase of the rhythms appeared to shift in the appropriate direction when studied under nychthemeral conditions whereas the endogenous oscillator either showed no consistent behaviour or, in the control of urate excretion, a shift in the wrong direction.8. The implications for people undergoing time shifts, in the course of shift work or transmeridional flights, are indicated.

No abstract

1. Seven solitary subjects, and two groups of four, spent from 5 to 13 days in an isolation unit without knowledge of time. Three solitary subjects and one group of four adopted fairly regular activity habits with a period of 25–27 h; one subject adopted a period of 30 h, and one of 27 h initially, decreasing to 24–25 h after a few days. One group of four awoke roughly every 24 h, after a sleep which was alternately about 8 h, or about 4 h and believed by the subjects to be an afternoon siesta. Two solitary subjects alternated sleeps of about 8 or 16 h, separated by 24 h of activity. 2. Deep temperature in all subjects oscillated with a period of 24–26 h, which was thus commonly distinct from their activity habits. 3. Urinary potassium followed a rhythm whose period, though usually close to, was sometimes distinct from, that of temperature. A secondary period corresponding to that of activity was also sometimes present. 4. Urinary sodium and chloride usually gave evidence of two periodic components, one corresponding to activity and the other to the rhythm of either temperature or of urinary potassium. 5. Urinary creatinine and phosphate usually followed the subject's routine of activity. 6. Plasma samples were collected on a few occasions and analysed for phosphate and 11–hydroxycorticosteroids. Changes in plasma phosphate were usually, but not always, associated with similar changes in urinary phosphate, and changes in plasma corticosteroids were often, but not always, associated with similar changes in urinary potassium shortly afterwards. 7. Observations are recorded on a subject alone in a cave for 127 days. His activity habits, though wildly variable, gave evidence of a period of 25·1 h and his urinary electrolyte excretion indicated a shorter period, of 24·6 h. During the following 3 days, when he remained in the cave but was visited frequently, his plasma corticosteroids and urinary potassium oscillated with a period of 16 h. 8. The possible mechanisms controlling these rhythms are discussed.

IDiurnal fluctuations of temperature and of sleepiness and wakefulness have been intensively studied (Kleitman, 1939), and have been shown to persist independently of any external rhythm for some time. The diurnal rhythm of urine flow has, however, received much less attention. With the recognition that,many variations of urine flow are due to varying secretion by the posterior pituitary, determined in turn by variations in hypothalamic activity, a reinvestigation of the existence or otherwise of an intrinsic rhythm in urinary flow appeared desirable.Many external factors which vary more or less regularly over 24 hr. affect urine flow, such as intake of food and fluid, exercise, sleep, external temperature. It was felt that 48 hr. was the shortest period in which a regular rhythm could satisfactorily be shown, and that the maintenance of absolute constancy of the external variables over this period would both be difficult and deter possible subjects. It was decided therefore to live on a 12-hr. cycle of meals, sleep, etc. and to observe any variations of urinary flow at corresponding times in successive cycles.Conditions were chosen in which it would be easy to reverse ordinary habits of waking and sleeping, and thus to find out how easily any inherent urinary rhythm might thereby be disturbed. METHODSThe author and five students were the subjects of the experiment. This was conducted in a remote and uninhabited shepherd's cottage on the island of Arran at about midsummer, which in 1950 coincided with a full moon. It was hoped thereby to minimize variations of temperature and illumination, as well as any possible psychological effect of contact with other people living on a 24-hr.cycle. On clear nights it was never completely dark but there were some hours of deep twilight, whilst on cloudy nights there were 2 or 3 hr. of darkness.For 48 hr. a 12-hr. cycle was repeated approximately as follows: 3-4.30, cook and eat main meal. 4.45-8.45, sleep. 9, eat a light snack. 9.30-11 or 11.30, gentle walk.

The existence and persistence of the habitual circadian (Halberg, Halberg, Barnum & Bittner, 1959) rhythms of renal function have been studied in Eskimos and other Arctic dwellers who are not subjected to the usual alternation of night and day, as well as in inhabitants of temperate latitudes spending the summer in the Arctic and living on abnormal time schedules (Lobban, 1960). When it was learnt that Mr G. Workman intended to spend 100 days in solitude in a pot-hole, the opportunity was taken to find out how his renal rhythm would respond to this complete isolation from any environmental rhythm, and almost complete isolation from contact with other human beings. METHODSWorkman entered Stump Cross Cavern, near Pateley Bridge, on 16 June, 1963 and re mained there until 29 September. His intention was to follow normal habits of sleep and waking, meals, activity and so on. He used a watch reading British Summer Time (B.S.T.), and had a field telephone with which he spoke to the owner of the cave, at the surface, once a day; apart from this, he had no contact with human beings. The preparation of meals, and other necessary chores involved in living alone, occupied a fair proportion of his time, and he was also actively engaged in further exploration of the cave system and clearing a blocked passage. No special effort was made to adhere to a uniform diet, but since all his food was taken down with him, his diet was somewhat monotonous; as assessed from 24 hr electrolyte excretions, it supplied potassium, 22-54 m-equiv, mean 37, and chloride, 80-250 m-equiv, mean 158; sodium was similar to chloride. The low intake of potassium resulted in an unusually high Na:K ratio in the urine. The temperature was constant, 700, and the atmosphere nearly saturated with water vapour. Light was provided mainly by candles, or by a miner's forehead lamp when he was exploring, but he used a Tilley lamp for short periods. Once a week, while he was underground, all the urine produced during two sleep periods and the intervening 'day' was collected, by spontaneous voiding at times convenient to the subject. When sleep was broken by the need to pass urine, the two samples were collected and analysed separately. On these days of urine collection he was usually, but not always, physically idle. The urine volume was read to the nearest 10, or occasionally 5 ml., and the time to the nearest 5 min, and a portion of each sample was preserved; these portions were left at a pre-arranged place for collection and despatch to Manchester, where they were analysed for chloride (Sanderson, 1952), sodium and potassium (flame photometer), phosphate (Fiske & Subbarow, 1925) and creatinine (Bonsnes & Taussky, 1945).When samples were received reasonably fresh, pH was measured by a glass electrode, and if