the basis of available photographic information, these surface currents cannot even be extended to the closest coracles. Although the reason remains unknown, failures under Similar conditions in the future can probably be prevented by using heavier mooring cable; cable diameters as large as '4 inch can be used without greatly increasing horizontal drag forces. During Shot Umbrella, none of the moors failed; however, seven of the close-in coracles overturned shortly after zero time. The precise reason for capsizing also remains a subject for speculation. Device~generated water waves do not appear Sufficient as a Single agent. The force on a coracle in a horizontal attitude, due to the 100-knot water-laden winds observed near surface zero, was estimated to be about 160 pounds. Increasing the water burden from an assumed 3 to 100 gm/‘ft® of air and increasing the sail area to that of a coracie rolled 30° to the horizontal raises this force to about 680 pounds. A force of this magnitude, if properly applied, is capable of overturning a coracle. Without experimental measurements of metacentric shift at high angles of inclination, a precise approach to this problem is not possible. It seems unlikely, however, that the proper conditions for coracle overturn due to base surge forces alone could have been reached. Perhaps, such base surge forces operating in conjunction with the turbulent water conditions existing inside the foam ring could have been sufficient for capsizing. tf, in spite of the previous discussion, device-generated water waves are assumedto be the reason for capSizing, their arrival times at the closer project stations have been calculated using data presented by SIO in Reference 74 and are given in Table 2.7 for comparison with the estimated time of overturn. Sometimes the time of overturn may be estimated from the GITR record (Section 3.3.2}. In about half the cases a sharp decline in the peak dose rate is roughly coincident with the calculated arrival of the device-generated waves, but the later GITR record does not appear to be that of an overturned coracle. Sharp decreases in dose rate could also be the result of a rapidly transiting radioactive cloud. Inspection of the preliminary photographic information appears to supportthis latter hypothesis. The estimated times of coracle overturn have been arrived at through a careful comparison of all gamma records along a given line of radial expansion and all currently available photographic information. In any event the estimated times of overturn appear to be so late as to preclude action of base surge drag forces. Ay 75