1912 CE, C. B. Davenport, in The American Naturalist, The Science Press; Volume XLVI, №. 543, page #134:Granting our inability to reason about genotypical constitution from the phænotypical, still, other things being equal, and in the long run and with great numbers of individuals an extremely high variant is more apt to belong to a genotype with a high mean than to one with a low or intermediate mean.