is slightly higher than the lowest collection but these occurrences are insignificant in view of the interpolated nature of the curves. The effective wind speed for each platform is the weighted mean of the wind speeds based on fallout amounts. Approximate particle densities and some estimated particle sizes from other studies are also listed.4»> ‘The agreement between the observed and computed bias directions is‘to be noted. Vector analysis revealed that the YAG 39-C, YAG 40-D and YAG 39-D experienced s.w. bias and their higher bias ratios agree with this. Their ground values should be determinable from s.w. platform-ground relationships as discussed. ‘The lower bias ratios of the m.w. systems are due to the inherent reduction of bias effects by winds of different direction. ‘The low bias ratio of YAG 40-c is in full accord with its low bias fraction. As mentioned, m.w. equivalent ground value determination must await further platform-ground information. Some indirect information concerning the equivalent ground value of the YAG collections has been obtained by water sampling; however, correlation between Platform and water sampling is not attempted in this study but is discussed elsewhere.3,10 Analysis of IST and Barge Data In addition to the normal air flow disturbances, the IST and barge platforms, because of their low positions, probably experienced other wind disturbances. Withthe winds impinging about the vessel's bow and sides the resulting updrafts and flow distortions could produce other bias effects. In the case of the stern platform, which was approximately 14 ft above the preceding top deck, the deck expanse and obstruction of the front platform contributed their share of flow irregularities with frontal winds. The resultant effect of these flow conditions is to complicate the normal bias and such complications cannot be defined at the present time. In order to present some indications of the bias situation, collection studies were performed (where possible) in the same manner as for the YAG platforms. The results are listed in Table 3; but it must be emphasized that these bias characteristics are, at best, rough approximations. In some cases it was exceedingly difficult to plot the collection curves and in others it was altogether impossible. The relationship between the two platforms of the YFNB 29 cannot be determined at this time. of the LST and barge biases were not attempted. Further studies CONCLUSION The variation of fallout collection within the standard platforms has been correlated with the air flow characteristics above the platform. In the case of multi~winds, sampling bias was further correlated tothe variapility in wind direction and associated fallout amount by a vector system. 19