On pairs of triangular numbers whose product is a perfect square and pairs of intervals of successive integers with equal sums of squares
2025
Read On pairs of triangular numbers whose product is a perfect square and pairs of intervals of successive integers with equal sums of squares (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-06598.txt.
Introduction
Vladimir Gurvich and Mariya Naumova wrote a 2025 math paper on triangular numbers whose product is a perfect square. Euler long ago described which triangular numbers are squares on their own. Pairs of such numbers clearly multiply to a square. But many other pairs work too, and those are the subject here.
The authors give clear formulas that describe all such pairs, using whole-number polynomials. That result lets them find all pairs of integer blocks with equal sums of squares, for block lengths that differ by one. A known link to near-isosceles right triangles falls out as a special case. The formulas are the paper's main new step.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-06598
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On pairs of triangular numbers whose product is a perfect square and pairs of intervals of successive integers with equal sums of squares Abstract A number N N is a triangular number if it can be written as N = t ( t + 1 ) / 2 N=t(t+1)/2 for some nonnegative integer number t t . A triangular number N N is called square if it is a perfect square, that is, N = d 2 N=d^{2} for some integer number d d . Square triangular numbers were characterized by Euler in 1778 and are in one-to-one correspondence with the so-called near-isosceles Pythagorean triples ( k , k + 1 , l ) (k,k+1,l) , where k 2 + ( k + 1 ) 2 = l 2 k^{2}+(k+1)^{2}=l^{2} . A quadratic number is the product Π = Π ( k , j ) = k ( k + 1 ) ( k + j ) ( k + j + 1 ) \Pi=\Pi(k,j)=k(k+1)(k+j)(k+j+1) for some nonnegative integer numbers k k and j j . By definition, it is the product of two triangular numbers and 4. Quadratic number Π \Pi and the corresponding pair ( k , j ) (k,j) are called square if Π \Pi is a perfect square. Clearly, ( k , j ) (k,j) is square if both triangular numbers k ( k + 1 ) / 2 k(k+1)/2 and ( k + j ) ( k + j + 1 ) / 2 (k+j)(k+j+1)/2 are perfect squares. Yet, there exist infinitely many other square quadratic numbers. We construct polynomials j i ( k ) j_{i}(k) of degree i = 2 ℓ i=2\ell and i = 2 ℓ + 1 i=2\ell+1 with positive integer coefficients a t a_{t} and b d b_{d} satisfying equations: k + j 2 ℓ ( k ) + 1 = k [ a ℓ k ℓ + ⋯ + a 1 k + a 0 ] 2 + 1 = ( k + 1 ) [ b ℓ k ℓ + ⋯ + b 1 k + b 0 ] 2 k+j_{2\ell}(k)+1=k[a_{\ell}k^{\ell}+\dots+a_{1}k+a_{0}]^{2}+1=(k+1)[b_{\ell}k^{\ell}+\dots+b_{1}k+b_{0}]^{2} and k + j 2 ℓ + 1 ( k ) + 1 = k ( k + 1 ) [ a ℓ k ℓ + ⋯ + a 1 k + a 0 ] 2 + 1 = [ b ℓ + 1 k ℓ + 1 + b ℓ k ℓ + ⋯ + b 1 k + b 0 ] 2 k+j_{2\ell+1}(k)+1=k(k+1)[a_{\ell}k^{\ell}+\dots+a_{1}k+a_{0}]^{2}+1=[b_{\ell+1}k^{\ell+1}+b_{\ell}k^{\ell}+\dots+b_{1}k+b_{0}]^{2} for positive integer a t , b d a_{t},b_{d} and ℓ \ell such that t = 0 , … , ℓ ; d = 0 , … , ℓ + 1 t=0,\ldots,\ell;\;d=0,\ldots,\ell+1 . We prove that all such pairs ( k , j i ( k ) ) (k,j_{i}(k)) are square and conjecture that there are no other, except the products of two square triangular numbers (of the same parity). We are also looking for pairs of intervals of successive integers with equal sums of squares. A simple case analysis shows that the problem is reduced to equation ( n − m ) 2 + ( n − m + 1 ) 2 + ⋯ + ( n + k ) 2 = ( n + k + j + 1 ) 2 + ( n + k + j + 2 ) 2 + ⋯ + ( n + 2 k + j ) 2 (n-m)^{2}+(n-m+1)^{2}+\dots+(n+k)^{2}=(n+k+j+1)^{2}+(n+k+j+2)^{2}+\dots+(n+2k+j)^{2} in integer j , m , k j,m,k and n n such that j ≥ 0 , m ≥ 0 j\geq 0,m\geq 0 and k > 0 k>0 . We rewrite it as a quadratic equation on n n and obtain an explicit formula for its discriminant D = D ( j , m , k ) D=D(j,m,k) . In case m = 0 m=0 we have D = Π ( k , j ) D=\Pi(k,j) . Hence, D D is a perfect square whenever Π ( k , j ) \Pi(k,j) is. Moreover, in this case both roots are integer. We also obtain some partial results for j = 0 j=0 . Cases m = j = 0 m=j=0 and k = 1 k=1 were studied in earlier works. SUBJECT CLASSIFICATION: Number theory, triangular and quadratic square numbers, (near-isosceles) Pythagorean triples, Pell’s equation. Vladimir Gurvich National Research University Higher School of Economics (HSE), Moscow, Russia, and RUTCOR, Rutgers,the State University of NJ, USA [email protected] , [email protected] Mariya Naumova Rutgers Business School, Rutgers University, Piscataway, NJ, United States [email protected] 1 Introduction Identities involving sums of squares of integers, usually consecutive, have been studied extensively for a long time; see for example, [ 1 , 13 ] . In this paper, we are looking for pairs of intervals of successive integers with equal sums of squares. Observation 1 . Without loss of generality (wlog) the problem is reduced to equation (n-m)^{2}+(n-m+1)^{2}+\dots+(n+k)^{2}=(n+k+j+1)^{2}+(n+k+j+2)^{2}+\dots+(n+2k+j)^{2} (1) in integer j,m,k and n such that j\geq 0,m\geq 0 , and k>0 . Proof. Equation ( 1 ) corresponds to the case of two disjoint intervals of lengths k+m+1 and k with distance j+1 between them. In general, two intervals might intersect. In this case, we just delete all terms of the intersection from both sides of the corresponding equality. Clearly, we obtain an equivalent reduced equality that corresponds to disjoint intervals. Furthermore, in this new equality j becomes equal to the size of the intersection plus 1. Furthermore, j=0 means that the two obtained disjoint integer intervals are adjacent, that is, at distance 1. Equation ( 1 ) assumes that (*) the left interval is longer than the right one. The difference of their lengths is m+1 , that is m\geq 0 and m=0 , the left is still longer by 1. In general, (*) might fail. Then we just negate all numbers in both intervals. Obviously, this transformation respects the equality and enforces (*). If both intervals are of the same length then m=-1 . Obviously, in this case ( 1 ) holds if and only if the considered two intervals are symmetric with respect to (wrt) 0, that is, one is the negation of the other. We do not consider this case, since it is trivial, in other words, we assume that m\geq 0 . ∎ Thus, we consider two intervals of k+m+1 and k integers at distance j+1 , where j\geq 0,m\geq 0, and k>0 are integers. In contrast, n may be a complex number, but we are interested only in integer n , which may be negative. We transform ( 1 ) to a quadratic equation an^{2}+bn+c=0 wrt n whose coefficients a,b,c and discriminant D=\sqrt{b^{2}-4ac} depend on j,m,k . Two solutions are given by formula n=\frac{1}{2a}(-b\pm\sqrt{D}) . If D<0 , both solutions are not real, if D\geq 0 both are real and they coincide if and only if D=0 . Both (or equivalently, each of them) are rational if and only if D=D(j,m,k) is a perfect square. Obviously, the denominator is a divisor of 2a . We