IN TRODU C TIONThere is a widely held view that dampness in the atmosphere has a large influence on how cold one feels when the temperature is low (between -40 ° and 0°C). This is surprising in view of the small solubility of water in air at low temperature and therefore the small effect water vapour has on rate of heat loss by conduction or radiation at these temperatures. Burton et al. 0955) showed that if anything dissolved water in the atmosphere should-reduce bodily heat loss, and in experiments• o on unclothed subjects they found that at temperatures of 9 and 14°C sensations of cold were significantly less at high humidities than at low, although skin temperatures were not affected by humidity. It has been suggested, though, that increased conductivity of clothing at high humidities at low temperatures could account for it (Winslow et al. , 1937).In order to investigate this effect subjective dampness was determined by choosing a panel of 6 subjects (two scientists, two technicians, one lab attendant and one tradesman) who commented on coldness and dampness after walking just under half a mile (800 m) in the open in winter in one of the colder parts of Eastern Canada. Their comments were compared with measurements of temperature and relative humidity (determined, as was the total water content of the air, from the frost point, using the Smithsonian Meteorological Tables). Also determined were wind speed and direction, cloud cover, and visibility of the sun. The observations were made once daily, at 08:15 hr, five days a week, over a period of 20 weeks between December and April 1966-1967.
ANALYSIS OF RESULTSThe panellists comments were scored one for very dry; two, dry; three medium; four, damp; and five, very damp. On occasions individual scores differed sharply, so they were averaged each day, and the average values, referred to herein as subjective dampness, were used for analysis.In Fig. 1 averaged scores are plotted versus relative humidity. There is considerable scatter in the results; the line drawn on the graph is that giving least squares deviation for the relative humidity values. The correlation coefficient for the points came to 0.47, indicating that, despite the scatter, there was a significant correlation between subjective dampness and humidity (the significance was tested by Students t test (see e.g. Moroney,1957, p. 311) which showed that the correlation could have arisen by chance with a probability of only about 0.1%). However, whether the sun was shining or not obviously influenced the results: crosses on the graph indicate overcast, and open circles indicate sunny conditions; the line subjective coldness = 3 almost completely separates the results for sunny from those