Let an=an−⌊an⌋, n=1,2,3,…. Since a2=a+(k−1)/4, it follows that ⌊a2n⌋=⌊an⌋+n(k−1)/4, and (a−1)⌊an⌋=(a−1)(an−an)=n(k−1)/4−(a−1)an, so, adding ⌊an⌋ to each side, a⌊an⌋=⌊an⌋+n(k−1)/4−(a−1)an=⌊a2n⌋−(a−1)an. Since a is irrational, the an form a dense subset of the open unit interval (0,1), so, by the preceding, the differences ⌊a2n⌋−a⌊an⌋=(a−1)an form a dense subset of the open interval (0,a−1). Finally, since ⌊a2n⌋−⌊a⌊an⌋⌋=⌊a2n⌋−a⌊an⌋⌋=⌊(a−1)an⌋, the conclusion follows.