Nature 431, 275–278 (2004)

There is an error in our data-reduction algorithm, which resulted in a systematic bias in the 27Al/24Mg ratios published in this Letter by a factor of 1.11. Consequently, the (26Al/27Al)0 values in the original Table 1 were underestimated (see corrected part of Table 1 below). The only significant change to our conclusions is that the initial 26Al/27Al at the time of formation of calcium–aluminium-rich inclusions (CAIs), as calculated from our data set, should be (5.83 ± 0.11) × 10-5, which is in agreement with that for CAIs from carbonaceous chondrites1. The main conclusions regarding the contemporaneous formation of some Allende chondrules and CAIs, as well as the brevity of the CAI-forming event, remain unchanged.

Table 1 Al-Mg isotopic data for Allende CAIs and chondrules