@icecolbeveridge ... Let me retract that halfway. Binomial is repeated Bernoulli tests, so I suppose the normal approx does follow from CLT