654 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
The average relative humidity for the subway for July and August 
was 57.5 per cent; for the outside air, 60.6 per cent; difference, 3.1 
per cent. 
The greatest average relative humidity occurred during the week 
when the average temperature was highest. During this period the 
relative humidity averaged 64.4 per cent. 
Percent of Humidity 
Fic. 3.—Weekly average relative humidity in the subway and streets from July 10 to 
November 13, 1905. These averages are made up of 47,456 observations. 
CHEMICAL CONDITION OF THE ATR. 
The chemical analyses of air were confined chiefly to determina- 
tions of carbon dioxide, for it was thought that no other test could 
give such a correct knowledge of the extent to which the air was 
vitiated by respiration, and none could be made on such a large scale 
as was wanted with so little probability of error. The method of 
analysis was accurate to within .03 part per 10,000. 
7 24 3i| 7 4 2 28) 4 tl 8 2 mime ac eo 
me 
YVolu 
¥ or 
(es ee a a ee eS eee ee 2 
—— Se Ae ed er 
BGRSOGGmh222 6p = eSeseea! 
Git Si Sai ee ee ee ere ee 
Fic. 4.—Weekly average carbon dioxide for the subway and streets from July 10 to 
December 25, 1905, including 1,772 determinations. 
b 
4 
Ports per 10,000 
About eighty samples of air were analyzed for oxygen. The differ- 
ence between the amount present in the subway and in the streets 
seemed so slight and uninstructive that the determinations were soon 
discontinued as a routine procedure. 
The carbon dioxide analyses produced results from which I de- 
rived the following conclusions: 
The average amount of carbon dioxide in the subway was a little 
larger than in the air of the streets. 
