Well i personally believe that:
This procedure can be done for all n. >!CorrectCorrect me if im wrong please.
Something we will need later:
T1: Two perfect squares differ by >!atleastatleast 3. Proof: x >=1: (x+1)^2-x^2= >!2x+12x+1 >= 3
Ok the base my proof is:
I will use induction on n. The >!inductioninduction base can be n=9 as in the >!questionquestion. In the induction step we >!knowknow that the hypothesis is true for >!n=kn=k-1 now we need to prove the >!hypothesishypothesis for n =k.
Ok now the proof begins:
Now we construct our 3 sets for >!n=kn=k-1. Now we name the sets :A,B,C. >!NowNow if the numbers 2 and 1 are in A >!andand we add k to A, and S(A) becomes >!aa square according to T1 if we take 2 >!oror 1 S(A) will be not a squre and if we >!addadd for exmaple 2 to another set and >!itit doesnt become a square then its ok >!otherwiseotherwise we dont pick 2 and we pick >!11. In the case which only one of 1, 2 >!areare in A(say x is that number), if when >!addingadding k to A, the sum becomes a >!squaresquare we can take x and put it in the >!setset where the other of 1,2 is if that set >!becomesbecomes a square we can just add >!thethe other number back to A. Now if >!bothboth of 1,2 are in the same set other >!thanthan A, if when adding k to A its sum >!becomesbecomes a square we can give 1 to A >!ifif the other set becomes a square too >!wewe can switch 1 and 2. Now the last >!casecase: if 1,2 are in separate sets other >!thanthan A, we can do the above >!proceduresprocedures for another set like B >!becausebecause this case ia covered. This >!solutionsolution (all of it) depends on T1.
I hope my answer is atleast partially true.