A proof of Dumont's conjecture. (Q2577014)

Let \[ \begin{aligned} r_{k}^{1(2)}(n) &:=\! | \{(x_1,x_2,\dots,x_k) \!\in \mathbb{N}^k : n = x_1^2 + x_2^2 + \cdots + x_k^2, \;\;x_i \equiv 1 \pmod{2} \}| ,\\ c_{k}^{1(4)}(n) &:=\! | \{(x_1,x_2,\dots,x_k) \!\in \mathbb{N}^k : n = x_1x_2 + x_2x_3 + \cdots + x_{k-1}x_k + x_kx_1, \;x_i \equiv 1 \pmod{4} \}| , \\ c_{k}^{3(4)}(n) &:=\! | \{(x_1,x_2,\dots,x_k) \!\in \mathbb{N}^k : n = x_1x_2 + x_2x_3 + \cdots + x_{k-1}x_k + x_kx_1, \;x_i \equiv 3 \pmod{4} \}| . \end{aligned} \] In unpublished work Dumont conjectured the identity \[ r_k^{1(2)}(n) = c_k^{1(4)}(n) - (-1)^kc_k^{3(4)}(n). \] The conjecture is proven in this paper using elementary manipulations of Lambert-type series, some \(q\)-series identities of Cauchy and Gauss, and a reformulation of the identity in terms of matrix algebra. As the author observes, several classical number-theoretic results concerning sums of squares and triangular numbers follow from this identity.