Unexplained periodic fluctuations in the decay rates of 32 Si and 226 Ra have been reported by groups at Brookhaven National Laboratory (32 Si), and at the Physikalisch-Technische-Bundesandstalt in Germany (226 Ra). We show from an analysis of the raw data in these experiments that the observed fluctuations are strongly correlated in time, not only with each other, but also with the distance between the Earth and the Sun. Some implications of these results are also discussed, including the suggestion that discrepancies in published half-life determinations for these and other nuclides may be attributable in part to differences in solar activity during the course of the various experiments, or to seasonal variations in fundamental constants. Following the discovery of radioactivity by Becquerel in 1896 [1] an intense effort was mounted to ascertain whether the decay rates of nuclides could be affected by external influences including temperature, pressure , chemical composition, concentration, and magnetic fields. By 1930, Rutherford, Chadwick, and Ellis [2, p. 167] concluded that "The rate of transformation of an element has been found to be a constant under all conditions." (For decays resulting from K-capture, or for beta-decays in strong ambient electromagnetic fields, the situation is slightly more complicated, since these decays are influenced by the electron wave functions which can be affected by external pressure or fields [3, 4, 5].) For 32 Si and 226 Ra, which decay by beta-and alpha-emission, respectively, fluctuations in the counting rates (in the absence of strong external electromagnetic fields) should thus be uncorrelated with any external time-dependent signal, as well as with each other. In what follows we show that neither of these expectations is realized in data we have analyzed for 32 Si and 226 Ra, thus suggesting that these decays are in fact being modulated by an external influence. Between 1982 and 1986, Alburger, et al. [6] measured the half-life of 32 Si at Brookhaven National Laboratory (BNL) via a direct measurement of the counting rate as a function of time. If N (t) denotes the number of surviving atoms starting from an initial population N 0 at t = 0, then the familiar exponential decay law, N (t) = N 0 e −λt , leads to ˙ N ≡ dN/dt = −λN 0 e −λt where λ = ln(2)/T 1/2. A plot of ln ˙ N (t) as a function of time is then a straight line whose slope is λ, which then gives the half-life T 1/2. At the time this experiment was initiated, the 32 Si half-life was believed to be in the range of 60 T 1/2 700 yr, and hence a multi-year counting experiment was needed to obtain a mea-sureable slope. As in other counting experiments, the counting rate for 32 Si was continually monitored in the same detector against a long-lived comparison standard, which in the BNL experiment was 36 Cl (T 1/2 =301,000 yr). Since the fractional change in the 36 Cl counting rate over the four year duration of the experiment was only O(10 −5), which was considerably smaller than the overall uncertainty of the final...
