This paper describes part of a research on the influence of environment on physiological reactions and thermal comfort. Experimental rooms and apparatus are described and the imperfections in some of the instrumental methods - especially in respect to the measurement of air change by tracer substances - are noted. A brief description of a method to measure air change is givenin which CO2 is used. The importance of limiting the rate of air change in rooms heated by fires - as a means to save heat - is stressed. Experiments were made with fresh-air ducts designed to draw air from the ceiling-roof space and deliver it into the flue without mixing the air of the room.