Protagoras comments on binomial variance problem