well-counter decay rates for the two types of particles deviate on both sides of the interval from 200 to 1,200 hours, and how the same curves fail to coincide, as they should for equivalent radio. nuclide compositions, when plotted in terms of 10‘ fissions. The latter shows the regions in which the primary radionuclide deficiencies exist. The previous considerations suggest that particles should be grouped according to type for the study of activity-size relationships. Figures 3.22 and 3.23 show the results of a study made in this way (Table B.9). A large num- ber af the particles for which size and activity data were obtained in the YAG 40 laboratory during Shots Zuni and Tewa were first grouped according to size (16 groups, about 32 microns wide, from 11 to 528 microns), then subdivided according to type (irregular or angular, spheroidal or spherical, and agglomerated) within each size group. The distribution of activities in each size group and subgroup was considered and it was found that, while no regular distribution was ap- parent for the size group, the subgroup tended toward normal distribution. Median activities were utilized for both, but maximum and minimum values for the overall size group were included in Table B.9 to show the relative spread. It will be observed that activity range and median activity both increase with size. Similar results for groups of particles removed from IC trays exposed aboard the YAG 39, LST 611, YFNB 13, and YFNB 29 during Shot Tewa are also included in Table B.9. These have not been plotted or used in the derivation of the final relationships, because the particles were removed from the trays and well-counted between 300 and 600 hours after the shot, and many were so near background that their activities were questionable. (This should not be interpreted to mean that the fallout contained a significant number of inactive particles. Nearly 100 percent of the particles observed in the YAG 40 laboratory during Shots Zuni and Tewa were active. } In the figures, the median activity of each size group from the two sets of YAG 40 data has been plotted against the mean diameter of the group for the particles as a whole and several of the particle type subgroups. Regression lines have been constructed, using a modified least- squares method with median activities weighted by group frequencies, and 95-percent-confidence bands are shown in every case. Agglomerated particles from Shot Zuni and spheroidal particles from Shot Tewa have not been treated because of the sparsity of the data. : It should also be noted that different measures of diameter have been utilized in the two cases. The particles from both shots were sized under a low-power microscope using eyepiece microm- eter disks; a series of sizing circles was used during Shot Zuni, leading to the diameter of the equivalent projected area D,, while a linear scale was used for Shot Tewa, giving simply the maximum particle diameter Dm. The first method was selected because it could be applied under the working conditions in the YAG 40 laboratory and easily related to the method described in Section 3.2.4 (Figure B.5); the second method was adopted so that more particles could be processed and an upper limit established for size in the developmentof activity-size relationships. The equations for the regression lines are given in the figures and summarized as follows: all particles, Shot Zuni, A « D,?‘, Shot Tewa, A < Dp’"*; irregular particles, Shot Zuni, A « Da?, Shot Tewa, A x D,'''; spheroidal particles, Shot Zuni, A « D,>"’; and agglomerated particles, Shot Tewa, A < D,,”"!. (Analogous relationships for Tewa particles from the YFNB 29 were derived on the basis of much more limited data in Reference 25, using maximum diameter as the measure of size. | These are listed below; error not attributable to the linear regression was estimated at about 200 percent for the first two cases and 400 percent for the last: all particles, A « D,,?""'; irregular particles, A « D,,'-; and spheroidal particles, A < Dy,**".) It may be observed that the activity of the irregular particles varies approximately as the square of the diameter. This is in good agreement with the findings in Reference 23; the radioautographs in Figures 3.14 and 3.17 show the activity to be concentrated largely on the surfaces of the irregular particles. The activity of the spheroidal particles, however, appears to vary as the third or fourth power of the diameter, which could mean either that it is a true function of particle volume or that it diffused into the molten particle in a region of higher activity concentration in the cloud. The thin-section radioautographs suggest the latter to be true, showing the activity to be distributed throughout the volume in some cases (Figure 3.16) but confined to 52