We appreciate Glikson's congratulations on our study of Early Archean spherule beds and we note that he accepts the central point of our paper—that the Cr isotopic data provide unequivocal evidence of extraterrestrial matter in three Barberton greenstone belt spherule beds. He also does not dispute our inference that these represent the distal ejecta of major large-body impacts. We agree that the simplest interpretation of the data for beds S3 and S4 is that these were basin-forming events, comparable to impacts during the Late Heavy Bombardment at 3.9 Ga that formed the lunar maria. In various papers (Lowe et al., 1989, 2003; Byerly and Lowe, 1994) we have suggested that the target was pre-existing mafic-ultramafic volcanics and that impacts might have triggered volcanism and reorganized tectonic regimes on a regional if not global scale. But this is a very different argument from the Green (1972) models for a fundamentally different tectonic style based on impacts for the Early to Middle Archean. We do not believe there are sufficient data to support such models.
While vapor fractionation may affect the chemistry of Archean spherule beds, we doubt that there is evidence that it is responsible for the platinum group element abundances in our samples. Our discussion of the Kyte et al. (1992) study of platinum group elements in bed S4 was meant to point out that the fractionated platinum group element pattern observed in these rocks was opposite to the effect expected by Au mineralization. Kyte et al. (1992) discuss in detail the relative partitioning of platinum group elements by magmatic and hydrothermal processes and by impact vapor fractionation, when considering the origin of the platinum group element abundances in bed S4. In rocks such as S4, where all the original minerals have been replaced, secondary processes such as diagenesis, hydrothermal activity, and meta-somatism are likely to have affected the platinum group element pattern much more than vapor fractionation. The small differences in condensation temperature between Pd and Au are unlikely to explain Pd/Au ratios in S4 that are 20 times CI chondrites. This is easily explained by mobilization of Au. We consider it likely that virtually all of the projectile will condense with the silicate fraction, resulting in very little platinum group element fractionation in the final ejecta deposit. Further, we find no evidence in Glikson's comment to support vapor fractionation. We note that the Pd/Ir ratios of published data on 2.56 Ga Hamersley Basin spherules are all greater than in chondrites, contrary to the assertion by Glikson. This is consistent with relatively high Pd concentrations (and Pd/Ir ratios) in crustal rocks. We cannot evaluate unpublished data on other deposits by Glikson and his colleagues.
We cannot agree with the assertion that extreme platinum group element enrichment is related to alteration of “iridium nanonuggets.” Kyte et al. (1992) reported that three micron-sized platinum group element (not just Ir) nuggets were found in a single pyrite grain in bed S4. They stated that such nuggets could not be the primary host of platinum group elements in that rock or others would have been found. There are no other published data on platinum group elements in any mineral phase in any spherule beds. Glikson implies that he has data on platinum group elements from minerals in S3, and we encourage him to publish these so that they can be properly evaluated.
Glikson suggests that because “winnowing and reworking of spherule-bearing beds by shallow-water currents can be expected to result in destruction of the delicate, originally glassy spherule structures, unless protected by rapid burial and subsidence, preservation of spherules in shallow-water environments is less likely.” It is noteworthy that a wide variety of seemingly fragile particles, including accretionary lapilli, pumice grains, and fluffy carbonaceous particles, were widely preserved in Barberton greenstone belt sediments deposited in environmental settings showing evidence of long-term wave and current activity. Rapid deposition, early silicification, and particles more robust than expectations have all played roles in the preservation of these materials. One might “expect” spherule destruction mainly if the “winnowing and reworking” involved prolonged exposure to currents and extended abrasion. However, fan deltas on a vegetation-free world would probably have been dominated by flashy events, like those in modern arid environments, and could have provided an excellent setting locally for the rapid burial and preservation of fall-deposited spherules. The sedimentology of clastic sediments associated with S3 and S4 in the Barberton greenstone belt has been extensively discussed (e.g., Nocita and Lowe, 1990; Lowe and Nocita, 1999) and a fandelta setting for Jay's Chert is reasonably well established. If Glikson has new sedimentological data from this unit that would support reinterpretation, he should present it. It seems unwarranted to question interpretations derived from extensive sedimentological data based on an untested model as to how someone thinks impacting should work. Perhaps it is the model and the associated expectations that should be reconsidered. These greenstone sequences also show dikes at virtually all levels containing a wide variety of debris derived from surface layers, including volcanic ash, carbonaceous matter, and impact debris. They do not bear on arguments for shallow- versus deep-water spherule deposition.