666 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
The largest number of bacteria found in a sample of subway dust 
was 2,000,000. 
FINAL CONCLUSIONS. 
A review of the results of the investigation warrants, in my opinion, 
the following brief statement of the most essential facts determined 
with respect to the quality of the air. | 
According to usual sanitary standards, based on chemical and bac- 
teriological analyses, the general air of the subway was always and 
everywhere satisfactory. The air in the cars in winter is not included 
in this statement. 
According to public opinion, based on the testimony of the senses, 
the air was everywhere unsatisfactory, especially during the summer 
months. 
My own conclusion was that the general air, although disagreeable, 
was not actually harmful, except, possibly, for the presence of iron 
dust. The strong drafts in winter at the stations and the lack of 
sanitary care exercised over the subway were, however, worthy of 
careful consideration in this connection. 
The high temperature of the subway was its most noticeably objec- 
tionable feature. Had it not been for the heat, it is probable that the 
other unpleasant features would have failed to arouse serious protest. 
The heat, as is well known, was due to the conversion of the electric 
power which ran the trains into friction. The amount of heat given 
off by the passengers was so small by comparison as to have had 
practically nothing to do with elevating the general temperature. 
The heat was most objectionable in the mornings and evenings of 
summer during the hours of greatest travel and when the air outside 
was cooler than during the rest of the day. 
The heat did not indicate that the air was vitiated or stagnant, as 
was popularly supposed. The subway was hot because a great deal of 
heat was produced in it, and stored by the materials of which the 
subway was built. That the heat did not escape rapidly enough for 
comfort was no proof that the air was not renewed often enough for 
health. 
The carbon dioxide and oxygen analyses indicated that the products 
of respiration were rapidly carried away. Among the 2,200 carbon 
dioxide determinations, most of which were made in the subway, no 
sample of air was found which contained above 8.89 parts of CO, per 
ten thousand volumes, and this amount was found under circum- 
stances which must be regarded as exceptional. 
The average excess of carbon dioxide in the subway over that in the 
streets, 1.14 parts per ten thousand volumes, showed that the air was 
renewed with remarkable frequency. In the absence of a census giv- 
ing the number of passengers in different parts of the subway at dif- 
