INTRODUCTION Relationships between the size and radioactivity of fallout particles play an important role in fallout phenomenology and model research. These relationships aid in the assessment of radiological situations and, more important, may reveal the manner in which activity is associated with the particles. Furthermore, the modes of association may provide some insight into the history of fallout material and suggest the mechan- ism of particle formation and contamination. It has been generally hypothesized that activity is a function of either particle surface area or volume. Experimental data to confirm this hypothesis is needed. In the case of coral fallout particles, radioautographic studies by Adamsl-4 have shown the existence of three general types of particles with intrinsic differences in activity association. Irregularly shaped particles generally have the activity concentrated on the exterior, indicating that it was deposited after the particle was formed. Spheroidal particles usually have the activity dispersed throughout, suggesting that it was incorporated during some molten state. Regard- ing the third type, particles with a dendrite-like structure, little is known of activity association. The existence of such differences neces~ sitates the consideration of particile shape in the activity-size relationships of coral particles. Other studies5»® have been limited to the composite activity associated with size fraction. However, certain deficiencies are inherent in this method of study. The approach does not take into account the variability of particle activity due to particle type. Moreover, particle size separation was generally accomplished by sedimentation, sieving and other egitating methods; since a substantial portion of coral particles are dendrite-like and fragile, these methods very likely cause particle breakup and subsequent errors. Regarded as smaller particles, the fragments cause inaccuracies in the data, particularly fragments from particles with activity concentrated on the exterior. It is essential, therefore, to employ a method of study that will consider particle type and will not break up fragile particles. Such requirements are met by the study of individual discrete particles.* This approach not only provides reliable data but also gives needed * %It is to be noted that Kikuchi, et. al. at examined a number of individual CASTLE particles, but neglected particle type; no correlation between size and activity was found. i

Select target paragraph3