planned end were adequate for the CaSTLE tests. As the operation progressed, however, Certain modifications were made as dictated by umusual circumstances which_arose. These modifications are discussed below. (1) Native populations: As indicated in the native populations chart (Tab G), the bulk of the indigenes within 500 NM cf GZ reside in the southeast quadrant. | Consequently, all atolls except UJELANG (and cNIWETOK for BIXINI shots) were considered to be in a favorable location with respect to fall-ov-. However, as indicated in Tabs K through P pertaining to each event, there wes a general movement of air particle trajectories to the east regardj.es:: of the initial directions of the winds near ground zerce. This phenomena was associated with all shots, the mid-level air particle trajectories turning toward the east as the cloud moved away from GZ, even to the point of taking place from positions relatively close to ground zero on the two shots (ROMEO and NECT..R) with pronounced southerly flow at ground zero. The net consequence of this action was the eventual transport of the most significant portion of the cloud (1.e., the mid-levels from approximtely 20,000 feet to 60,000 feet) to the east where subsequent subsidence of the debris could place a considerable amount in the trade wind flow to be brought back into the general area of the tests. The gross effect of this mechanisa (as well as other factors) was profoundly indicated on the BRAVO shot. effect was apparent to a mich lesser degree on the remaining shots. Its The diminished contamination of the remaining shots appears chargeable to more faverable initial wind conditions, the low yield of the one remaining land surface shot (XOON), and to the fact thet all subsequent high yield devices were shot fron barges. The pertinent facts and ideas which became epparent as the shot schedule progressed, and due to the phenomena described above, are as follows: em 29