1108 
the atmosphere of organisms. The comparatively large 
number of microorganisms found in rain water by various 
investigators is further evidence that at least a sizeable 
fraction of the air-borne organisms have been captured 
and removed from the atmosphere by falling raindrops. 
While it has not been definitely proved that micro- 
organisms ever act directly as condensation nuclei in the 
atmosphere, at least one investigator has noted that the 
rate of fall of spores in moist air is several times greater 
than in dry air, presumably because the organism has 
increased its mass through the absorption of water 
vapor. 
It would thus appear that the condensation and pre- 
cipitation of water vapor are the important mechanisms 
for removing microorganisms from the atmosphere. 
The transport to the surface through the effects of 
turbulence with subsequent settling or impingement 
appears to be the secondary process. This reasoning, 
of course, does not apply to the larger organisms whose 
free rates of fall are appreciable and which are sup- 
ported aloft only within strong convective currents or 
through their own flight activity. 
Vertical. and Horizontal Transport of Organisms. 
Turbulence in the atmosphere tends to create a uni- 
form distribution of properties throughout the air mass. 
This action, however, is of far greater magnitude than 
the mere diffusion of properties between layers; ap- 
parently the mixing takes place through the motions of 
eddies or discrete parcels of air which are pushed from 
their original surroundings at irregular intervals. Thus, 
it would be expected that air-borne microorganisms, 
whose source is at the surface of the earth, would be 
distributed throughout the entire depth of at least the 
troposphere,° but that the density of organisms would 
decrease the greater the distance from the source. 
Biological work to date shows this to be true. Proctor 
[21] notes a general decrease in both dust particles and 
bacteria with elevation and an increase in the bacterial 
count in the vicinity of wooded areas, while Rittenberg 
[26] has presented data which show a general decrease 
in the mold count over the oceans with increasing dis- 
tance from shore. It would appear that the horizontal 
distance over which the individual microorganism may 
be transported is almost limitless and is largely de- 
termined by its ability to survive the atmospheric 
environment.® The meteorologist, of course, is interested 
in the total distribution of microorganisms; the 
biologist, on the other hand, concerns himself only 
with what he terms the effective distribution. The ef- 
fective distribution is a function of the virulence and 
concentration of organisms and of the conditions bear- 
5. Living spores of several common molds were caught in a 
spore trap released from the balloon Explorer IT at 72,500 ft 
and set to close at 36,000 ft [27]. 
6. Living spores of various fungi have been caught from 
aeroplanes above the Caribbean Sea 600 miles from their 
nearest possible source [17], while allergens have been identi- 
fied at least 1500 miles from their probable origins. In spite of 
the usual low concentrations at the source, marine bacteria 
have been collected 80 miles inland from the nearest seacoast 
[40], 
BIOLOGICAL AND CHEMICAL METEOROLOGY 
ing upon their multiplication and spread at the place 
and time of deposition. 
As an example of the sometimes complex meteor- 
ological circumstances that must precede the effective 
long-distance dissemination of important plant patho- 
gens, Stakman and Christensen [82] cite the spread of 
stem rust on wheat through the North Central States 
during 1942 and 1943. In both years, barberry bushes 
(the secondary host) in the Virginias became heavily 
infected about May 15 because of favorable weather 
circumstances during the preceding period of develop- 
ment. Throughout the first half of June, southeast 
winds, accompanied by rains (favoring multiplication), 
carried the rust spores west-northwestward, causing 
heavy infection through much of Ohio and Indiana 
and as far as southern Michigan and northeastern 
Illinois. While the south-to-north air movement was 
not unusual, the destructive epidemics resulted only 
when all meteorological factors favorable for rust de- 
velopment operated in conjunction. 
Supporting evidence on the long-distance dissemi- 
nation of plant pathogens is given by data collected in 
Canada which show that urediniospores of leaf and stem 
rust of wheat are regularly introduced each spring into 
Manitoba by spores from the United States. These 
rusts do not overwinter in Canada, but the uredinio- 
spores are invariably found in the air in advance of the 
infection of the wheat crop. Of even greater interest is 
the fact that the rusts do not survive the hot summers 
of the south and every fall must be reintroduced by 
wind-borne spores from the north. The meteorologist 
cannot help but note how the maintenance of such a 
biological cycle is favored by the normal seasonal change 
in the general atmospheric circulation over these areas. 
Numerous similar examples of the long-distance dis- 
semination of organisms appear in the very extensive 
literature on aerobiology. A large fraction of the papers 
contain aerobiological data which should be further 
analyzed by the meteorologist. 
Although turbulence tends to create uniformity of 
properties throughout an air mass, at any given instant 
this air mass must contain variously large and small 
masses (eddies) with somewhat different properties from 
those of the surrounding air. This no doubt accounts 
for at least part of the extreme variability in bacterial 
counts within small areas, or over short intervals of 
time, which has been noted by various investigators. 
The more intense or prolonged the turbulent action, 
however, the more nearly will uniformity of properties 
be approached. At the same time, however, there must 
be a general decrease in the density of the property | 
relative to the total air mass as the distance between 
the observation and the source area increases. 
If this “uniformity of properties’? were actually at- 
tained in a given region (which would be the case were 
the process one of pure diffusion), the spatial distri- 
bution of microorganisms would approximate a fre- 
quency distribution of the Poisson type, and the ratio 
of the variance (7.e. mean squared deviation) of the 
counts to the mean of the counts made in the region 
should approximately equal 1.00. A higher ratio indi- 
