Depositional Histories of Sand Grains from Surface Textures

Abstract

BROWN¹ has drawn attention to the occurrence of features of “glacial origin” on sand grains from freshly weathered granite, from beaches, and from streams, all presumably lacking evidence of previous glacial history. We would like to clarify certain misleading statements presented by Brown¹, not with the purpose of discrediting her findings but rather to argue with the interpretation of her data with respect to previously published literature on this subject. Krinsley and Donahue's transmission microscope work² of 1968, referred to by Brown, has been greatly expanded on by much detailed work since the use of the scanning electron microscope to examine sand grain surface textures became widespread^3–9. The statement made by Brown, quoting Krinsley and Donahue² “that an assemblage of textures should characterize any one abrasive process, and that no single feature can be considered completely diagnostic,” was modified by Margolis and Kennett⁴ to read “… no single surface feature can itself be used to identify the origin of a sand grain, but rather that certain combinations of features are highly diagnostic when statistically significant percentages of the grains in the sample are examined”. With regard to Brown's statements on features of glacial origin Margolis and Kennett⁴ have written: “Features that in previous studies have been individually regarded as diagnostic of glacial origin such as large conchoidal fractures and breakage blocks, are also occasionally found on grains from high turbulence beaches and also occasionally on desert sands. When combinations of surface features are examined statistically, however, glacially derived sand grains exhibit unique combinations of characteristics”^3,6. The point here is that the technique had advanced beyond presenting individual photographs as evidence of the history of a sand, and now involves a more quantitative treatment of data.