BACTERIA OF THE AIR 703 
in gelatin. These reactions are regarded as satisfactory to establish 
the identity of B. coli. 
In some lal)oratories a direct plating of the sample of water in lactose- 
litmus agar or upon Endo medium is practised, but this procedure is 
considerably less sensitive than the fermentation enrichment method 
outlined above. 
Colon bacilli may occasionally be isolated from considerable volumes 
of water— 10 or 100 cc— when they cannot be detected with regu- 
larity in 1 cc. or less. Very little significance attaches to such results, 
because experience has shown that even springs in uninhabited regions 
may occasionally contain a few colon bacilli, derived probably from 
chance contamination with the feces of wild animals. If, on the 
contrary, colon bacilli are regularly present in a water supply to such 
an extent that a cubic centimeter of the water gives a positive culture 
in a decided majority of attempts, that water is viewed with suspicion. 
If the organism is regularly present in 0.1 cc, the water is judged unfit 
or dangerous for human consumption until it is purified. 
Other organisms have from time to time been jjroposed as indi- 
cators of pollution— thus streptococci and gas bacilli ha\'e been studied 
in this connection— but up to the present time they have not been 
accepted as authoritative criteria for evaluating the potability of 
water supplies. 
BACTERIA OF THE AIR. 
Bacteria when dried and attached to dust particles may be wafted 
into the air and remain suspended there for considerable periods of 
time. Even the gentlest air currents suffice to prevent their settling 
out. At high altitudes and over large bodies of water the bacterial 
population of the air is very small indeed; over large cities and cul- 
tivated land the number of organisms in the air is frequently much 
greater. Heavy rains and snow^ tend to remove bacteria from the 
atmosphere, while dry windy weather increases the aerial contamina- 
tion. 
Usually the more hardy organisms alone are found in the air, but in 
houses and hospitals pathogenic bacteria may be detected occasionally; 
probably the extrusion of minute droplets of sputum^ containing these 
organisms is a most potent factor in air contamination by bacteria. 
Several methods have been proposed for the estimation of the num- 
ber of bacteria in the air; that of Winslow,- which consists essentially 
is aspirating a definite volume of air through two flasks, each of which 
contains melted nutrient gelatin, is the simplest and most direct. 
Comparatively little has been accomplished thus far from a quantita- 
tive study of the bacterial pojnilation of the air; it is possible that 
an attempt to isolate specific types of pathogenic bacteria from theaters 
and other places where large numbers of people meet might throw 
some light upon certain features of the air transmission of bacterial 
infections which are not well understood at the present time. 
1 See Droplet Infection, p. 91. ■ Science, 1908, 38, 28. 
