1. Let d=(2012t+1,2013t+1) then it is easy to see that d=1. Thus, both 2012t+1 and 2013t+1 are perfect square if and only if (2012t+1)(2013t+1)=y2 for some positive integer y. We have
(2012t+1)(2013t+1)=y2⇔4⋅20122⋅20132t2+4⋅2012⋅2013⋅4025t+4⋅2012⋅2013=4⋅2012⋅2013⋅y2⇔(2⋅2012⋅2013t+4025)2−1=4⋅2012⋅2013⋅y2
Putting x=2⋅2012⋅2013t+4025 we have the equation x2−4⋅2012⋅2013y2=1.
It is clear that 4⋅2012⋅2013 is not perfect square, so the Pell equation of 1st type has infinitely many solutions. The fundamental solution of this equation is (x,y)=(4025,1), thus all its solutions are given by formula
{x0=1,x1=4025,xn+2=8050xn+1−xn,y0=1,y1=1,yn+2=8050yn+1−ynn≥0.
By induction, we can prove that x2i+1 is congruent to 4025(mod2⋅2012⋅2013) for all i and each value 2⋅2012⋅2013x2i+1−4025 gives us a positive integer t satisfying the required condition.
So there exists infinitely many positive integers t such that t both 2012t+1 and 2013t+1 are perfect squares. (Q.E.D)
2. Let d=(mn+1,mn+n+1) then
d∣(mn+n+1−mn−1) or d∣n, therefore d∣(mn+1−mn) or d∣1.
Thus d=1 or in other words the numbers mn+1,(m+1)n+1 are relatively prime.
So, both mn+1 and (m+1)n+1 are perfect square if and only if
(mn+1)((m+1)n+1) is perfect square.
Assume that (mn+1)((m+1)n+1)=y2 for y∈Z+. We have
m(m+1)n2+(2m+1)n+1=y2⇔4m2(m+1)2n2+4m(m+1)(2m+1)n+4m(m+1)=4m(m+1)y2⇔(2m(m+1)n+(2m+1))2−1=4m(m+1)y2
Putting x=2m(m+1)n+(2m+1) then we have following equation
x2−4m(m+1)y2=1(∗)
This is the Pell equation of 1st type. Since 4m(m+1) is not perfect square then (*) has infinitely many solutions.
The fundamental solution of equation (*) is (x,y)=(2m+1,1), so all its solutions (xi,yi) can be written in the form
{x0=1,x1=2m+1,xi+2=2(2m+1)xi+1−xi,y0=0,y1=1,yi+2=2(2m+1)yi+1−yii≥0
By induction, we will prove that x2i is congruent to 1 modulo 2m(m+1) and x2i+1 is congruent 2m+1 modulo 2m(m+1) for all i=0,1,2,... (***)
Indeed,
- For i=0, by the recurrent formula of (xi) we see that (*) is true.
- Assume that (*) is true for i, i.e x2i is congruent to 1 and x2i+1 is congruent 2m+1 modulo 2m(m+1). We have
x2i+2x2i+3=2(2m+1)x2i+1−x2i≡2(2m+1)(2m+1)−1=8m(m+1)+1≡1(mod2m(m+1)),=2(2m+1)x2i+2−x2i+1≡2(2m+1)−(2m+1)≡2m+1(mod2m(m+1)).
Thus, (***) is true for i+1.
By induction principle, (***) is true for all i.
Next, we will establish the recurrent formula for ri=x2i+1 where i=0,1,2,....
We have
ri+2=x2i+5=2(2m+1)x2i+4−x2i+3=2(2m+1)(2(2m+1)x2i+3−x2i+2)−x2i+3=(4(2m+1)2−1)x2i+3−2(2m+1)x2i+2=(4(2m+1)2−1)x2i+3−(x2i+3+x2i+1)=(4(2m+1)2−2)x2i+3−x2i+1=(4(2m+1)2−2)ri+1−ri
Putting ri=2m(m+1)si+(2m+1) then the sequence (si),i>0 is well-defined and consists of positive integers.
Putting in recurrent formula of (ri), we get
2m(m+1)si+2+(2m+1)⇔2m(m+1)si+2⇔si+2=(4(2m+1)2−2)(2m(m+1)si+1+(2m+1))−(2m(m+1)si+(2m+1))=2m(m+1)(4(2m+1)2−2)si+1−2m(m+1)si+4(2m+1)(2m+1)2−1=(4(2m+1)2−2)si+1−si+8(2m+1)
We can compute r0=x1=2m+1 so s0=0 and
r1=x3=2(2m+1)(2(2m+1)2−1)−(2m+1)=16m(m+1)(2m+1)+2m+1,
so we have s1=8(2m+1).
We have the recurrent formula for (si) are
{s0=0,s1=8(2m+1),si+2=(4(2m+1)2−2)si+1−si+8(2m+1),i≥0
From which we can see that all terms of (si) are divisible by 8(2m+1).
Moreover, by putting x=2m(m+1)n+(2m+1) we can easily see that n satisfies the required condition if and only if n=si,i=1,2,3,... (note that s0=0 is not a positive integer).
Thus, all value of n is divisible by 8(2m+1). (Q.E.D).