Mark Buchanan’s otherwise admirable piece on random matrices (10 April, p 28) failed to cite the remarkable Tracy-Widom distributions, which, when applied to a given random matrix, will give the ...