This thesis concerns an investigation of the ability of sheep to withstand cold. It includes studies of the initial cold resistance, the ability to acclimatize to cold, and of the variation within and between breeds of sheep. The cold resistance of shorn Scottish Blackface, Southdown and Welsh Mountain sheep, equal numbers on high or maintenance levels of nutrition, was measured by the rate of decline of rectal temperature under acute cold exposure (-20°C; 4 m.p.h. wind). Their ability to acclimatize to cold was determined by the effect of three types of cold exposure on subsequent cold resistance. These were; 2 weeks at a moderately subcritical temperature (+8°C), up to 8 hrs. acute exposure, or a combination of the two. Associated physiological responses such as changes in skin temperature, heart rate, respiration rate and skinfold thickness were also measured. Cold resistance was greatest in the Blackface and least in the Welsh sheep. The mean rates of decline of rectal temperature were 0.6}6, 1.105 and 1.791 °C/100 min. exposure for Blackface, Southdown and Welsh sheep respectively. High plane sheep had greater cold resistance than low plane, but there was no evidence for interactions between breed and nutrition. There was considerable variation between individuals within breeds; coefficients of variation ranged from 25% to 66%. Individual repeatability of performance was high. The cold resistance of all breeds was increased by previous exposure to +8°C, while an increase in cold resistance as a result of acute cold exposure alone was clearly shown by the Blackface sheep. These findings indicated that sheep could acclimatize to cold. Blackface sheep had a much greater capacity for acclimatization than either Southdown or Welsh sheep. The mean rates of decline of rectal temperature were 74%, 19% and 32% slower respectively after acclimatization. In general both high and low plane sheep showed similar amounts of acclimatization. There was some inconclusive evidence for breed x nutritional interactions in ability to acclimatize. It was concluded from indirect evidence and by analogy with previous work on rodents that acclimatization probably resulted from an enhanced capacity of the sheep to maintain high rates of metabolism under cold exposure, while in Blackface sheep a slight change in the site and efficiency of heat production involving a reduction in shivering thermogenesis may also have contributed. Prolonged exposure to +8°C caused an elevation in rectal and skin temperatures and in heart rate when subsequently measured at a thermoneutral temperature. Rectal temperatures were, on average, 0.15 C higher, ear and foot temperatures 3.2 C and 2.5 C higher respectively, while heart rates increased by 40-50%. This evidence implied a considerable increase in the basal metabolic rate. Vasomotor responses under subsequent acute cold exposures were delayed. Vasoconstriction occurred, on average, 4°C and 7°C later respectively in the ears and feet. Acute cold exposure appeared to have a similar effect on basal metabolism. Sheep kept at moderately sub-oritical temperatures for long periods of time allowed, on average, a reduction in body temperature of 0.4 C. This was also shown during initial stages of acute cold exposure. The increase in cold resistance and the associated responses appeared to have slightly different properties. Both were induced by prolonged exposure to +8 C or by acute cold exposure, but apparently only the increased resistance to body cooling was able to persist without continuous low temperature stimulation. Evidence for acclimatization was also found in new born lambs. Lambs born and reared for 2 weeks at 0°C maintained a rate of fall of rectal temperature 30% slower during acute cold exposure than comparable lambs which had been kept at +25 C. The relevance of these findings to sheep in their natural environment and to climatic studies in the laboratory was discussed. The vasomotor responses of Blackface, Southdown and Welsh sheep under mild cold exposure in full fleece were also compared. In general, estimates of the critical temperature, defined as the ambient temperature obtaining immediately after vasoconstriction of both extremities, were in good agreement with those of other workers who used calorimetric determinations. These were +1 C, +3 C and +7 C for the Blackface, Southdown and 7'elsh sheep respectively. Estimates for high plane 3heep were generally 2°C lower than those for comparable low plane sheep. Southdown sheep, by comparison with the two hill breeds, appeared to have a rather low critical temperature relative to their fleece length, wool weight per unit area and bodyweight. This was thought to be due to their more comprehensive fleece cover on the face, ears and feet. It was suggested that vasomotor responses may be useful indices of variation in critical temperature within breeds but their validity in comparisons between breeds seems questionable when the sheep are unshorn.