The temperature readings are somewhat unsatisfactory because they usually include only the maximum and minimum records for each day. Minimum temperatures are found to be perceptibly higher in the cities than in the country, partly because of city heat, but chiefly because the smoke blankets prevent the escape of the heat at night. In Pittsburgh the minimum temperature averages about 4° higher than the temperature in the adjacent stations (Kimball). In the day time the upper surface of the smoke clouds absorbs portions of the sun’s heat, and this inevitably lessens the heat on the ground surface. The solar energy has been reduced 40% by smoke clouds (12), and it is possible that very little heat penetrates to the earth’s surface during a very dense smoke cloud. In London the surface temperature during a given reading was 44° F., while 59 ft. directly above it was 51.5°, a difference of 7.5°. Between 9 A. M. and 3 P. M. the surface temperature rose 7.5°, while the roof reading rose 16.5°, or more than twice as much. Although thoroughly satisfactory data are not available it is undoubted that smoke (which is most abundant in the atmosphere in the winter time, owing to lessened convection) lowers very perceptibly the diurnal winter temperature in smoke-producing cities. Finally, smoke indirectly affects the electrical potential of the atmosphere. I know of no direct study of this question, but it is possible that smoke causes a decrease in the potential because it tends to increase humidity. An increase of humidity decreases the potential (save possibly during fogs, as noted later, 14). We are also told that the potential increases with the turbidity (Trubung) of the atmosphere, and consequently with high barometric pressure. Since smoke tends to decrease the diurnal temperature, particularly in the winter time, it apparently tends to increase the diurnal potential, as the potential decreases with increasing temperature (14). Having thus discussed the atmospheric changes caused by smoke we now turn to the second question. How do the various smoke-induced weather states affect the mental and bodily health and happiness of human beings? It is evident that the problem can be formulated largely in terms of sunshine dynamics. On the one hand, we have the psychic and physiological influences of positive states of sunshine: brightness, luminosity, warmth, dryness, the presence of the infra-red, spectral and ultra-violet rays, and high electrical potential (because of dryness). On the other hand, we have the effects of the negative states: gloom, darkness, cloudiness, fogginess, rain, moisture, cold and low electrical potential. It is utterly impossible to separate all of these components, and it will, therefore, be no easy task to measure the influences of each component.