The published version of the original paper contained a Sr isotope analysis of Hornblende-pyroxene quartz monzodiorite sample H25, from the shoshonitic part of the Cambrian Yzerfontein pluton on the West Coast in South Africa. The value of 87Sr/86Sr, calculated at our newly determined U–Pb zircon age of 535 Ma, was given as 0.70885 ± 2 (2σ). In the paper, this was reported as an anomalously high value and we hypothesised that this might have been due to the crustal component of this particular magma having been somewhat more radiogenic than for the rest of the shoshonitic samples from the Yzerfontein pluton. We emphasise that the isotopic and elemental analyses were carried out on the same solutions. Nothing in the Rb/Sr ratio indicated accidental contamination of the sample, and nothing in the isotope results suggested analytical problems. However, this result was sufficiently surprising that we provided AEON Labs (UCT) with a second aliquot to dissolve and analyse. The redetermined value of 87Sr/86Sr535Ma is 0.70506 ± 2, which is very similar to all the other determinations for the shoshonitic rocks of the pluton. Although we remain unable to explain the original result, we believe it to have been erroneous and that the correct initial Sr isotope ratio of sample H25 is 0.70506. Accordingly, we present a revised Table 1, showing the Sr and Nd isotope data, with the correction, and a revised version of Figs. 6 and 13 in the original paper, as Figs. 1 and 2 here, with the correct value for H25 plotted. This amendment has no impact on the main conclusions of the original paper.