1. The large size of the project. If more-limited objectives had been adopted, and the measurements to accomplish these objectives allotted to several smaller projects, the amountof field administrative work and the length of time key personnel were required to spend in the field could probably have been reduced. In future tests, the total number of shot participations should be kept to the minimum compatible with specific data requirements. 2. The difficulty of maintaining adequate communications between the test site and NRDL. Despite arrangements to expedite dispatches, frequent informal letters, and messages transmitted by sample couriers, several cases occurred where important information was delayed in transit. 3. The use of instruments developed by other projects. Malfunctions were frequent in such cases but were probably due partly to lack of complete familiarity with the design of the instrument. This is the principal reason why the water-sampling results are incomplete and of un- certain reliability. 4. The operational characteristics of certain project instruments. The time-of-arrival de- tectors (TOAD) were developed for the operation and had not been proof-tested in the field. They tended to give good results when located on stable stations, such as barges or islands, and poor results when located on stations like the skiffs. It seems probable that minor design modifica- tions would suffice to make this a dependable instrument. The honeycomb inserts used in the open-close total collector (OCC) exhibited a tendency to spall and should be modified for future use. The sizes of the collecting areas of the always-open collector, Type 2 (AOC,), and incre- mental collector (IC) should be increased if possible. Complete redesign of the gamma timeintensity recorder (TIR) to improve its response characteristics, reduce its size, and makeit a self-contained unit was obviously required for future work and was initiated during the field phase. 5. The commitments of the project to supply early evaluations of field data. Because of the nature of fallout studies, inferences drawn from unreduced data may be misleading. Despite the urgency associated with studies of this kind, interim project reports should be confined to presenting the results of specific field measurements. 5.1.2 Technical. The generai conclusions given below are grouped by subject and presented for the most part in the same order that the subjects are discussed in the preceding chapters. In a sense, the values tabulated and plotted in the text constitute the detailed conclusions, because they represent the numerical results derived from the reduced data of the appendixes. For this reason, numerical values will be extracted from the text only if some generality is evident or to illustrate an observed range. Although the conclusions presented are not neces- Sarily those of the authors whose works have been referenced in the text, interpretations are usually compatible. Buildup Characteristics. 1. The time from fallout arrival to peak radiation rate was approximately equal to the time of arrival for all stations and shots. Activity-arrival rate was roughly proportional to mass- arrival rate for the solid-particle shots, Zuni and Tewa. A similar result was obtained for outlying stations during Shot Flathead, although this proportionality did not hold for Shot Navajo nor for the close-in collections from Shot Flathead. 2. The shape of the activity-arrival-rate curve was not markedly different for solid- and Slurry-particle shots. In both types of events, the time from the onsetof fallout to the time when the radiation rate peaked was usually much shorter than the time required for the remainder of the fallout to be deposited. There was sometendency for slurry fallout to be more protracted and less concentrated in a single major arrival wave; however, statistical fluctuations due to low concentrations of particles and small collector areas were responsible for most of the rapid changes observed after the time of peak. Where fallout concentrations were sufficiently high, good time correlation was ordinarily obtained between peak rate of arrival and peak radiation rate. 3. Particle-size distributions varied continuously with time at each station during the solid- Particle shots, activity arrival waves being characterized by sharp increases in the concentra151