In an effort to detect a pattern, the five regions of Table 1 pertaining to slope were assigned the following approximate time-spans in samples related d in the rels of | Samples. for the as ost of the ficiently expected slopes of rion that entire region. ible 1 the and 108e Because the points of inflection of the various curves do not coincide, these limits for the five periods had to be arbitrary. (curves 9-104 Omitting the first en- st region inclusive), the slope for a curve was entered in the blank gpace so that a slope was available for each of the five time-periods involved. the five 8, <100, 100-300, 300-700, 700-1200, and 1200-2300. eight curves, and, therefore, considering only the post-Mike material 10 Table slope days, respectively: Computation of mean and standard deviation (VSx2/(n-1)), not standard error, of the 92-96 slopes for each of the five time-periods gave, respectively, these results: -1.35%0.45; -1.26+0.28; -1.25t0. 34; -1. 25t0. 37, -1.78t0.79. The mean slope of -1.35 for the period up to 100 days is steeper than for the three following periods, -1,26, -1. 25, and -1.25, but is less steep than for the last period -1.78. Likewise, the standard deviation of slopes was greater for the first than for the succeeding three periods, but less than for the fifth, and this same variability is also discernible from the graphs. There is thus evidence of a general increase in rate of decay of post-Mike samples during the last three years, over the first three years, with the exception of certain Samples to be discussed later. The present decay rates may be compared to the theoretical gross beta decay curve for slow-neutron-induced fission products