648 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
ples near the ground was it possible to avoid attracting curious 
crowds of persons whose presence would have rendered the samples 
valueless. Tests made of air from different elevations indicated that 
no substantial error was made in taking samples near the pavement. 
Very few samples of air were taken in the cars. Persons familiar 
with the conditions of crowding in the cars of the New York subway 
at practically all hours of the day will appreciate the inconvenience 
with which delicate and bulky scientific apparatus could be used 
among persons standing as close together as it was physically pos- 
sible to stand. Furthermore, the question at issue was not whether 
the passengers in the cars obtained good air or not, but whether the 
air outside the cars was satisfactory. 
The part of the road which was in operation during the period of 
this investigation extended from the lower end of Manhattan Jsland 
northward to Ninety-sixth street and Broadway, where it divided, 
one branch continuing along Broadway to One hundred and fifty- 
seventh street, and the other eastward and northward until it crossed 
under the Harlem River and reached that part of the city known as 
the Bronx. 
Nearly all of this road was underground. There was a short, ex- 
posed portion of a few blocks covering a valley at One hundred and 
twenty-fifth street, and the branch to the Bronx, after crossing the 
Harlem, soon emerged upon an elevated structure, which it did not 
leave to the end of the line; but the parts of the subway which were 
not underground were not considered in this investigation. 
The length of the road, about 21 miles, and the rather wide variety 
of conditions which occurred in it made it desirable to confine the 
investigation as far as practicable to a representative section. 
There was no difficulty in selecting this section. The road between 
Ninety-sixth street and the Brooklyn Bridge was, in every respect, 
the most important. Further on in this paper it will be shown that 
this section was divisible into two parts, distinct differences both as to 
details of construction and the condition of the air being noticeable 
between the part north of Fifty-ninth street and that south. 
Nearly all the studies recorded in this paper, except those of tem- 
perature and humidity, refer especially to the representative section 
between Ninety-sixth street and the bridge. In many cases, however, 
they have a much wider application. 
The length of the section was about 6 miles. The cubic air space 
included was, in round figures, 26,100,000 cubic feet, including the 
stations. 
The section was four tracks wide, excepting a piece of tunnel which 
ran between Forty-second street and Thirty-fourth street. Here 
there were two tunnels of two tracks each, running side by side, cut. 
through the rock. 
