Petrography and Diagenesis of Cherts From Lake Magadi, Kenya

Abstract

Pleistocene cherts from alkaline Lake Magadi, Kenya, are believed to have formed inorganically from sodium silicate precursors (magadiite--NaSi 7 O 13 (OH) 3 .3H 2 O and/or kenyaite--NaSi 11 O (sub 20.5) (OH) 4 .3H 2 O). They have been cited as modern analogs for cherts found in both ancient lacustrine deposits and Precambrian iron-formations, but few criteria exist to distinguish Magadi-type chert from other cherts in the geological record. We studied approximately twenty chert samples and one magadiite sample from Lake Magadi to develop such criteria and to shed light on the mechanism of the magadiite-quartz conversion. Characteristics of the type Magadi cherts are 1) a groundmass mosaic of fine quartz crystals that vary in orientation from random to rectilinear; 2) finely disseminated inclusions of silicate clays, zeolites, and/or carbonates (predominantly calcite); 3) large crystal molds (probably after trona) concentrated near sample margins; and 4) inward-directed shrinkage cracks and/or more irregular internal voids filled with chalcedony, silicate clays, zeolites, and/or carbonates (predominantly calcite). Shrinkage cracks, which define surface reticulation patterns, are frequently cited as evidence of a Magadi-type origin for ancient lacustrine cherts, but they are not unique to alkaline lake cherts. Based on a reconnaissance of ancient cherts of various origins, no single characteristic appears to be unique to Magadi-type chert, with the possible exception of the rectilinear or grid-work orientation of the quartz crystals. This texture appears to be inherited from the precursor magadiite, which displays a similar extinction pattern due to the presence of 10-20 mu m spherical aggregates of plate-like crystals. This pseudomorphing implies direct volume-for-volume replacement of magadiite by quartz. Void-filling chalcedony and the high densities of the cherts, however, indicate that silica was added during and after the conversion from magadiite to chert. Petrographic observations suggest the magadiite-quartz conversion takes place in multiple stages, with early conversion in brines and later leaching by more dilute meteoric water. A reconnaissance of 35 chert samples from 16 different formations, ranging in age from Pleistocene to Archean, yielded only two samples with convincing grid-work fabrics, both of which had already been interpreted as Magadi-type chert on the basis of other characteristics. This suggests that the distinctive textures which characterize the type Magadi cherts may be useful for recognizing other inorganic alkaline lake cherts in the stratigraphic record.