Solution:
Let us first prove the case n=1. Knowing that ⌊λ2⌋,⌊λ3⌋, and ⌊λ4⌋ are perfect squares, we want to show that ⌊λ⌋ is a perfect square. Setting a2=⌊λ2⌋, we have the inequalities:
a2≤λ2<a2+1
from which, in particular, λ2<(a+1)2, hence a≤λ<a+1 and ⌊λ⌋=a. Squaring the previous chain of inequalities, we obtain
a4≤λ4<(a2+1)2
and, since a4 and (a2+1)2 are consecutive perfect squares, ⌊λ4⌋=a4. Consequently λ4−a4<1. Now, using the hypothesis λ≥1,
λ3−a3=(λ−a)(λ2+λa+a2)≤(λ−a)(λ(λ2+λa+a2)+a3)=λ4−a4
so we also have
a3≤λ3=a3+λ3−a3≤a3+λ4−a4<a3+1
which implies ⌊λ3⌋=a3. By hypothesis, we therefore know that a3 is a perfect square, hence a=⌊λ⌋ is also a perfect square, as required.
Let us now prove the general case, by induction on n. The base case, n=1, has been established. For the inductive step, suppose we know that, if ⌊λn+1⌋,⌊λn+2⌋,…,⌊λ4n⌋ are all perfect squares, then ⌊λ⌋ is a perfect square. From the fact that ⌊λ(n+1)+1⌋,⌊λ(n+1)+2⌋,…,⌊λ4(n+1)⌋ are all perfect squares, we want to deduce that ⌊λ⌋ is a perfect square. If we also knew that ⌊λn+1⌋ is a perfect square, then we could apply the inductive hypothesis and conclude immediately. Now, setting λˉ=λn+1, in particular, ⌊λˉ2⌋,⌊λˉ3⌋, and ⌊λˉ4⌋ are three of our perfect squares. From the case n=1 applied to λˉ it therefore follows that ⌊λn+1⌋ is a perfect square, and this concludes the proof.