mer often exceeding 90° F., and minimum temperatures in winter frequently going below 0° F. San Francisco, on the other hand, has a semi-tropical climate, with temperatures of 90° F. or over occurring but two or three times in a year, and minimum temperatures below 40° F. being equally rare. In addition, the climates of the two cities differ greatly in respect to the amount and duration of sunshine, cloudiness, rainfall, relative humidity, wind velocity and direction, and the various other elements which constitute climate. The mean annual temperature is therefore an inadequate indication of climatic conditions, and can not alone serve as a basis of comparison.
Too much emphasis is also placed upon the temperature itself—our feeling of comfort is by no means entirely dependent upon the reading of the dry-bulb thermometer. An ideal curve of comfort might show but littleto the thermograph trace. Relative humidity is so important a contributory factor that the wet-bulb rather than the dry-bulb thermometer is often the better indicator. The feeling produced by a temperature of 100° F. experienced in southern Arizona is wholly unlike that accompanying a similar temperature in an eastern city, the difference being due primarily to the marked difference in relative humidity. Other factors also affect one's feeling of comfort, such as sunshine, wind velocity, barometric pressure, and atmospheric electricity. Some time it may be possible to give correct relative weights to each of these factors in determining their effect upon the man in perfect health. While temperature doubtless will receive the greatest weight, the other factors are by no means negligible.
Night air is occasionally referred to as though it is different from day air, and convalescents are sometimes urged to avoid it as dangerous. While there are obvious physical differences between night air and day air there is little diurnal change in chemical composition. Atmospheric air is a physical mixture which when perfectly dry consists principally of nitrogen, oxygen, argon, and carbon dioxide, in which the relative proportions remain fairly constant, and in which the first two named constitute more than 99 per cent. by volume. Up to heights greater than the summits of the highest mountains the percentage of oxygen, an element necessary in the respiration of both plants and animals, shows no appreciable variation. Carbon dioxide, however, which forms but .03 per cent. by volume, or .05 per cent. by weight, of the air, shows both an annual and a diurnal variation. By volume it is 23 per cent. greater in summer than in winter, and is 12 per cent. greater at night than during the day. Since carbon dioxide does not become dangerous until it constitutes considerably more than 1 per cent. of the air we breathe, the change from day to night can not account for the supposedly offensive feature of night air. Water vapor, which never exceeds 4 per cent. by volume of atmospheric air, is important as far as