AND ITS INHABITANTS 165 



myself among the number, have always supposed that dry 

 weather is the best. There is no questioning the diagrams, 

 however, for, as has already been said, they are based on 

 millions of deaths. Moreover, the same features appear in 

 each of the five areas into which the United States is divided, 

 and in France and Italy when taken separately as well as 

 together. 



One of the most striking features of the diagrams is that 

 although their general form is alike and although they all 

 show substantially the same optimum, the lines are far apart 

 in Figure 31, the diagram for the northeastern United States, 

 nearer together in Figure 32 for France and Italy, and nearest 

 of all in Figure 33 for California. This means that while 

 people in all these regions are strongest and most energetic at 

 essentially the same temperature and humidity, a departure 

 from these apparently ideal conditions does much more harm 

 in France and Italy than in the northeastern United States, and 

 still more in California. We might add that between the 

 northern and southern parts of France and Italy there is the 

 same sort of difference as between Figures 31 and 32, or 32 

 and 33. This at first sight seems puzzling. The climates that 

 are generally considered best for invalids actually show 

 greater extremes in the death rates than do the more rugged 

 northern climates. This happens not only in southern Cali- 

 fornia where there are many invalids, but in San Francisco 

 and southern Italy where there is no influx of such people. 



The explanation is apparently found in the variability of 

 the different climates. Measurements of factory work and 

 studies of the marks of students show that changes of tempera- 

 ture, especially a moderate drop, are distinctly stimulating. 

 Other evidence also indicates that variability, both from season 

 to season and from day to day, is much better than uniformity. 

 Apparently people who live in variable climates acquire a 

 power of resistance unknown among those where relative 



