128 
BIOLOGY: E. HUNTINGTON 
TABLE 1 
Mean Temperature and Deaths in New York City, 1900-1911 
MEAN TEMPERATTIRE DEATHS PER DAY 
MONTH 
Actual 
Smoothed* 
Actual 
Smoothed* 
January 
30.2 
(30.2) 
256 
(256) 
February 
30.7 
31.5 
266 
257 
34.4 
34.3 
238 
253 
March 
37.5 
38.4 
270 
248 
44.0 
43.4 
215 
242 
April 
48.1 
48.9 
267 
240 
55.5 
54.6 
210 
231 
May. 
59.3 
60.2 
239 
227 
66.5 
65.2 
218 
225 
June 
68.5 
68.9 
224 
226 
August 
72.2 
71.6 
236 
241 
July 
73.5 
(73.5) 
266 
(266) 
* The smoothing is done by the formula ^- = b 
entire population. This alters the apparent optimum, but not more 
than a degree or two, for relatively few people are away in September 
and June. Therefore, on the basis of the death-rate the optimum 
temperature in New York City appears to be 63° or 64°F. 
Table 2 shows the optimum temperature for three groups of Ameri- 
can cities and for a group of foreign countries, all being reckoned in 
the same way as New York. 
Aside from New Orleans the optima range from 58° at San Francisco 
to 70° at St. Louis. Local differences arise from humidity, epidemics, 
the season at which vacations are taken, and other minor causes. Never- 
theless the averages for the four groups differ surprisingly little, the 
range being only from 62.7° to 65.8°. This seems to indicate that 
whether we are dealing with northwestern Europe, the northern, cen- 
tral, or southern United States, or an Asiatic country such as Japan, 
deaths are most numerous at about the same temperature. When al- 
lowance is made for the effect of summer vacations in the American 
cities, it appears that the optimum is probably about 63° F. Since we 
are dealing with the mean temperature of day and night, such an opti- 
mum means that man is physically at his best when the temperature 
rises to about 70° at noon, but not when it stays all day at that level. 
(2) Daily work. People's daily work is perhaps the best test of 
their physical condition. During the past few years I have collected 
statistics of the work of piece-workers in factories from Connecticut 
to Florida. A comparison of their work with the mean out-of-door 
temperature gives the results shown in table 3. 
