AIR CLEANSING BY CONVECTIVE STORMS
069
remove insoluble particles. The filtrate was reduced by evaporation
and the residue ultimately transferred to a stainless-steel planchet
and counted for radioactivity. The filters weredried, placed into identi-
cal planchets, and similarly counted. Early in the program the samples
were counted only for gross beta activity by using a low-background
(1 count/min) beta counter. After the resumption of atmospheric testing
by the U.S.S.R. in September 1961, radioactivity levels rose sufficiently
to permit determination of individual radionuclides by gamma spectrometry.
In each rainwater sample fission-product radionuclides were de-
termined in three groups: (1) “!ce, “4ce—pr; (2)Ru, 6Ru—!°8Rh;
and (3) *Zr—*Nb. In addition, “°Ba-—‘“°La were determined when fall-
out was fresh. The necessity for grouping the radionuclides as indicated arises from the method of analysis.
In the investigation of the mechanisms whereby fission products
are brought to earth by rain, the changes of radioactivity levels with
time provide the best clues. Thus the precision of the analysis proce-
dure is more important than its accuracy. The precision of sample
preparation and counting is estimated to be approximately 15 to 20%.
This
is
entirely
acceptable
in view of the magnitude of observed
sample-to-sample variations.
It will be seen later that time variations of the several radionuclide groups are largely parallel. Nevertheless, it was felt that random
errors would be minimized by computing the total of the measured
radionuclides, Therefore, unless specified to the contrary, only the
variations of the total are considered in the discussion. For con-venience, the concentration of total measured radionuclides is some-
times referred to as the “concentration.”
Pollen Analysis*
Insoluble material was separated from each sample by centrifuga-
tion. A known amount of a readily identifiable tracer pollen, which
would not normally be found in rain and a small amount of glycerine
jelly were thoroughly mixed with the insoluble residue, Several microscope slides were prepared from each sample. Each slide was made by
placing a single drop of the jellied mixture ona slide and covering it
with a glass cover slip. The number of grains of each pollen species of
interest and the tracer pollen were counted under a microscope for one
or more slides from each sample. Under the assumption that equal
fractions of the total number of tracer andnatural pollens were present
on the slides counted, it was possible to compute the total number of
*An alternate procedure to the one described here was used for the rain of
Sept. 1, 1961, and has been described by Gatz and Dingle.‘