Workers in factories and a postal sorting office were questioned concerning their subjective thermal sensations, and the replies assessed according to a scale of standard sensations. Measurements of the thermal environment were made concurrently. The investigation was confined to workers engaged in light or sedentary activity during summer months in southern England.

The upper limit of the comfort zone is determined in terms of the temperatures at which more than 20% of people questioned experienced thermal discomfort.

Discontinuities in the linear relationships between thermal sensation, described on a numerical scale, and the thermal indices of the environment are shown to occur under conditions which suggest the onset of sweating beneath the clothing. The comfort zone is also determined in terms of these critical temperatures.

The regression constants relating thermal sensation to thermal conditions are determined, and the optimum conditions for comfort are deduced. The accuracy of prediction of thermal sensation from thermal measurements is examined.

The importance of clothing in requirements for thermal comfort is illustrated in two ways. Postal workers wearing regulation uniform were found to require conditions cooler than those required by factory workers wearing clothing of their own choosing. Male and female factory workers were found to require thermal conditions not significantly different, this being attributed to their selection of clothing appropriate to their personal thermal requirements.

I am indebted to Prof. G. P. Crowden for granting facilities for this research and for encouragement throughout. Dr T. C. Angus and Dr T. Bedford gave valuable advice, and Mr P. J. Williams rendered technical assistance. The co-operation of the managements and workers at the factories was greatly appreciated.