At the Nevada Test Site, these various collectors were placed on the ground or on low platforms. In the Pacific, various seaborne craft were also used for the location of such samplers. Thus, one has used Liberty ships outfitted as instrumented test vessels traversing fallout areas, instrumented landing craft modified to operate in a fallout zone, and unmanned barges, anchored in the path of prevailing winds. Clouds were almost exclusively sampled by means of filter units mounted in pods attached to aircraft. The samplers, having a fixed geometry, were designed to approach isokinetic conditions as much as possible, but deviations from such conditions occurred as a result of variations in the altitude of sampling (barometric pressure), air velocity (air speed of the sampling aircraft), and ambient temperature. Several designs have been used. Originally the filters were ACC Type V (Army Chemical Corps), fabricated from an asbestos base. Dissolution of these filters was quite cumbersome, and their impurity content was unsatisfactory. Hence, the transition was made to cellulose-based IPC-1478 filter paper impregnated with "Kronisol" (di-butoxyethylphthalate). The filtration and permeability characteristics of this material have been investigated by several researchers (Inst. of Paper Chemistry, 1960; Stafford and Ettinger, 1970a, 1970b, 1971). Exact duplication of field sampling conditions tn the laboratory is very difficult, and most studies only approximated field conditions. Diffusion ts the primary mechanism of particle collection, particularly for the smaller-size particles. For the larger particles, the relationship between pore size and particle size dominates. Thus, the efficiency is dependent upon the particle size. For particles larger than a few microns, the efficiency approaches 100%, decreases to about 20% at 10-2 - 10-! um, and may then increase again towards smaller particle sizes. As a result, measured size distributions will usually be biased against the smaller sizes. Sticky wires were an important sampling method used in the Roller Coaster project conducted at the Tonopah Test Range to simulate accidental explosions of nuclear weapons under storage conditions. Arrays of wires coated with a sticky substance were suspended in the anticipated path of the cloud from tethered balloons (Traceriab, Division of LFE, 1966). Collection occurred by inertial impaction and clearly discriminated against the smaller particles. Nevertheless, this method yielded estimates of the activity distributions in the clouds projected on a vertical plane approximately perpendicular to the wind direction. In addition to the gross activity measurements on the wires, the particles in the samples could also be further analyzed. On the other hand, besides the limitations imposed by the collection efficiency of the wires, the sticky-wire method for collecting cloud samples is limited in altitude and by some lack of mobility of the system resulting in possible poor cloud interception as a result of wind direction changes. Turning now to the "miscellaneous" category, we mention the following: High-volume samplers (hi-vols) were used primarily in the peripheral areas of the test site and outside the test site mostly to determine atmospheric levels of radioactivity for monitoring purposes. Scoops were (and are) used to sample soil for subsequent analyses. To be mentioned here is the "Weasel," occasionally used in the fifties, being a remotely operated M-29 tank equipped with a scoop. Surface samples have also been obtained by coring, to obtain depth profiles of radioactivity. In recent years, special samplers have been designed and used by Anspaugh et al, (1974) to determine resuspension on the Nevada Test Site. These samplers consist of a vertical array of collection tubes positioned at a slant and are used passively. Finally, radar has been used to obtain cloud profiles, particularly at early times, and to estimate parcicle size distributions in clouds. ANALYTICAL DATA Since the inception of nuclear testing, a large number of properties has been measured often on isolated debris particles, but also on integral samples consisting of debris, nonradioactive admixtures to the clouds such as soil, and contaminants, usually windblown dust. Detailed analyses have been performed on samples from some selected tests, but some data are available from most shots. Limitations in the depth of the analyses were imposed by the large requirements of manpower, equipment, and time. Nevertheless, the detailed work that has been done encompasses a wide range of yields and detonation conditions. Table 2 lists the number of tests for which various data are available in greater or lesser detail. Particle isolation was accomplished mostly by location of the individual particles by autoradiographic techniques followed by micromanipulation. These particles were usually about 2 um or more in diameter. However, not all radioactive particles are identified, because autoradiography is limited by some minimum level of activity for each length of exposure. The submicrometer radioactive-particle population is not represented at all, and the remainder of the population is only partially represented. The properties that have been measured are listed in Table 3. are further discussed below. These measurements PARTICLE SIZE DISTRIBUTION Size distributions have been measured primarily to estimate debris transport and for input into fallout models. Together with radtoactivity data, they are used to evaluate the hazards to the biosphere (including man) resulting from local, intermediate, and worldwide fallout. They have been determined by means of optical or electron microscopy (or both) of integral samples and of