AIR OF NEW YORK SUBWAY—SOPER. 659 
Although it seemed likely from these reasons that most of the bac- 
teria in the air of the subway were derived from the streets, there 
was ground for concluding that some, and among them objectionable 
kinds, were due directly to the presence of the people. It is practi- 
cally certain when great crowds are packed together, as they often 
were in some stations and most cars, that dangerous bacteria are, at 
least occasionally, transmitted from person to person. An obvious 
feature of this danger lies in the fact that people talk, cough, and 
sneeze into one another’s faces at extremely short range under such 
circumstances. 
The numbers of bacteria in the air of the subway varied with the 
amount of travel. They were most numerous when the trains were 
most numerous, and fewest when the trains were fewest. 
When the trains were blocked many of the bacteria disappeared 
from the air. In one case the bacteria were reduced from 1,800 to 
250 in about an hour in this way. 
The effect of sweeping the platforms with brooms, without first 
taking precautions against raising dust, was noted. On one occasion 
the numbers of bacteria were increased by sweeping from about 5,000 
to 13,000, and remained above 8,000 for at least three-quarters of an 
hour—the time covered by the observation. 
It was not found that any harmful germs were capable of multiply- 
ing in the oil which dripped from the machinery of the cars upon the 
broken stone ballast and wooden ties of the roadbed. 
The lubricating oil apparently removed and collected from the air 
large numbers of bacteria, many of which soon ceased to exist. 
The pneumococcus was found capable of retaining its virulence in 
dried sputum in the subway for twenty-three days. This is in marked 
contrast to the findings of Wood, who reported that the pneumococcus 
was killed in four hours in sunlight. 
With few exceptions, there were not so many bacteria in the air of 
the toilet rooms as in the rest of the subway. In some cases the num- 
bers were much greater. 
The proprietary disinfectants used in the toilet rooms had no germ- 
icidal or deodorizing value. Furthermore, they produced counter 
odors of a peculiarly unpleasant character. 
The numbers of bacteria recovered from the dust of the subway 
averaged 500,000 per gram. 
The largest number of bacteria found in subway dust was 2,000,000 
per gram. Still greater numbers probably could have been found by 
selecting the specimens of dust toward this end. 
For comparison with the numbers of bacteria found in dust from 
the subway, it is interesting to note that dust which had accumulated 
under similar circumstances in a Broadway theater contained 270,000 
