Covered and naked barleys from the Himalaya differ significantly from each other in a multivariate way (Murphy and Witcombe 1986). Two hypotheses which could account for this phenomenon are (1) that the caryopsis covering gene (or a gene closely linked with it) exerts a wide-ranging pleiotropic effect, and (2) that the covered and naked types have evolved to differ with respect to the alleles fixed at many loci. Experimental evidence is presented which supports the latter, an explanation which has implications for our understanding of barley phylogeny and evolution.