Solution
The problem requirement is equivalent to the proof: When dividing the set {n, n + 1,...,2n) into two subsets, there exists a two-element subset whose sum is a perfect square.
Let x be the largest positive integer such that x2 ≤ n. Then we have n < (x + l)2. ie n ≤ x2 + 2x. Since n ≥ 100 so (x + l)2 > 100, ie x > 9. From here, we have
2n - (2x2 - 2) ≥ 2x2 - (2x2 - 2) > 0,
(2x2 - 2) - (2x2 - 4x + 3) = 4x - 5 > 0,
(2x2 -4x + 3)- (2x2 - 8x + 6) = 4x - 3 > 0.
(2x2 - 8x + 6)-n ≥ (2x2 - 8x + 6)- (x2 + 2x) = x(x - 10) + 6 > 0.
I guess
n < 2x2 - 8x + 6 < 2x2 - 4x + 3 < 2x2 - 2 <2 n.
Thus, three numbers 2x2-8x+6, 2x2-4x+3 and 2x2-2 are three numbers in the set A ={n, n + 1,...,2n).
When the set A is divided into two subsets, two of these three numbers will belong to the same subset. Which (2x2 - 8x + 6) + (2x2 - 4x + 3) = (2x - 3)2, (2x2 - 8x + 6) + (2x2 - 2) = (2x - 2)2 and (2x2 - 4x + 3) + (2x2 - 2) = (2x -1)2 are all perfect squares, so we have something to prove.
|