48 WORLDWIDE EFFECTS OF ATOMIC WEAPONS portions of the pilot study, should be flexible and should be allowed to change during the study, as the first results obtained will serve as a useful guide in further sampling. Thus, experiencewill tell whether the sampling program outlined in this section is adequate. An important part of the sampling program is the ‘‘mechanistic”’ sampling, i.e., sampling aimed at determining the physical form of the contaminant and the methods of nonlocal dissemination. This sampling program should becarried out in conjunction with what might be termed the “biological” sampling. The latter is carried out to determine how and to what extent the material is absorbed by the biosphere. We shall describe the former program only briefly; it is the most difficult and will take the longest time to develop. If data are to be obtained on an urgent basis (partially because of preater contamination after the CASTLE test series), it is believed that the results of the biological sampling. program will be the most significant. The first “mechanistic” problem is to determine the particle size and nature of particles. (See Chapter 2 and Appendix If.) There are strong indications that much of the particulate matter important to long-range considerations is in the form of ultra-finely-divided glasslike particles, possibly chemically inert. Sampling of the atmosphere requires the handling of matter in this new form. There is so little experience with aerosols involving particles of the ranges probably occurring that the problem of sampling the atmosphereis one of no minor difficulty. There does, fortunately, seem to be two pos- sible solutions to it—electrostatic precipitation and diffusion barriers. It appears that the electrostatic precipitation process, being extremely efficient for the collection of very-small-diameter particles, will be adequate for the task. Failing this means, the use of a diffusion barrier does seem to be promising (see “Analysis for Strontium 90” on page 58). Galelladecay of the pre- cursor, the chemically nonactive isotope Kr°°, there is a strong posstbility that the contaminant will be available in nearly atomic form to plate out on the newly formed particles and atomic debris. If so, the Sr°° should be easily available for solution. This hypothesis could be tested by various experiments designed to determine rate of ‘weathering’ of debris after SAMPLING AND RESEARCH PROGRAM 49 deposit. How rapidly plants take up Sr°* may also provide a measure of the degree of plating. Another important aspect of the “mechanistic” program is to determine the scavenging and other methods of bringing down the contaminant from the lower and upper atmosphere. Upto altitudes obtainable by aircraft the atmospheric sampling program appearsrelatively simple. The great question of what to do for samples at heights higher than those obtainable by aircraft has not been answered completely. There are possibilities of using balloons for indicating the general radioactive levels by telemetering or by actual collection of a sample. The need for these observations stems from the lack of knowledge of atmospheric processes. It has been suggested, and the suggestion appears reasonable, that fine particulate matter may be held for long petiods in the stable regions of the stratosphere. Such a storage would lead to a slow, relatively steady deposition as the debris was admixed with the weather layers and brought down by the processes of eddy diffusion, fall, and rain scavenging. Some evidence of a small but measurable “drip” of debris has been reported by Merril Eisenbud of the AEC New York Operations Office. It should be noted, however, that there is very little knowl- edge as to how the troposphere is cleaned of the debris or of how much deposited debris is raised from the ground by wind and redeposited. Until an adequate series of samples is taken from the atmosphere, it will be impossible to show, in an unambiguous fashion, how much debris is retained in the atmosphere and how rapidlyit is removed. A first requirement of the “biological” sampling program is to establish the usefulness and sensitivity of the analytical procedures for Srthat are described later in this chapter. Concurrent with this step it is necessary to determine the general levels of Sr® that may be anticipated in the samples that will be used. This is needed both to set the sensitivity for the analytical procedures and to determine the size of samples to be collected. Such preliminary information can probably be obtained from severaltypes Of $a : ' The first of these samples are some of the fallout specimens collected by the AEC monitoring system. (Do they contain Sr®°, and how much? A recent preliminary analysis showed that the samples do contain radioactive strontium, but no distinction between the 8&9 and 90 isotopes was