“…Note that two distinct classes of observed rates are displayed, corresponding to different contributions from the two d,u hyperfi ne states: (a) Experiments (8a, 14, 54, 55) separate the two hyperfine molecular formation rates X��I" (b) Although hyperfi ne populations are not always known exactly fo r experiments (26,52,53,56,57), they mainly correspond to the steady-state situation and measure an average molecular formation rate Xddl' that is a superposition of molecular fo rmation rates Zdw Within this data set there seems to be a normalization problem in the pioneering Dubna experiment (52), which obtains a similar T dependence but has a rate lower by a fa ctor of '" 4 than all subsequent experiments. Bubble chamber results (58) are not shown in Figure 5, but if interpreted properly…”