Primitive lattice points in rational ellipses and related arithmetic functions (Q1123923)

Let \(Q(u,v):=au^ 2+buv+cv^ 2\) be a positive definite quadratic form with integer coefficients a, b, c and \(D:=b^ 2-4ac<0\). For \(n\in {\mathbb{N}}\) let \(r_ Q(n):=\#\{(\ell,m)\in {\mathbb{Z}}^ 2:\) \(Q(\ell,m)=n\}\), \(\rho_ Q(n)\) the number of such pairs with \(\gcd (\ell,m)=1\) and for \(0<x\in {\mathbb{R}}\) let \(A_ Q(x)\), \(B_ Q(x)\) denote the corresponding summatory functions, respectively. Concerning the `rational' ellipse problem it is known that \(A_ Q(x)=Cx+O(x^{\alpha})\), \(C:=C(D):=2\pi | D|^{-1/2}\), still hold at least for any \(a>7/22\); see the forthcoming paper of \textit{M. N. Huxley} [Proc. Lond. Math. Soc., III. Ser. (to appear; Zbl 0659.10057)]. In the primitive case, assuming the validity of the Riemann hypothesis (RH), using \(\rho_ Q(n)=\sum_{d^ 2| n}\mu (d) r_ Q(n/d^ 2)\) and the complex integration method, the author proves (Theorem 1) \[ B_ Q(x)=(6C/\pi^ 2)x+O_ Q(x^{(1- \alpha)/(3-4\alpha)+\epsilon}) \] for any \(\epsilon >0\) and supersedes the result given by \textit{B. Z. Moroz} [Monatsh. Math. 99, 37-42 (1985; Zbl 0551.10038)]. Relating an extension of this result to a large class of planar domains see \textit{W. Müller} and the author [Lattice points in planar domains: Applications of Huxley's `Discrete Hardy-Littlewood method' (to appear in Springer Lecture Notes Math. (1990))]. Furthermore the author states (without proof, but assuming RH) a more general result for the summatory function of the arithmetical function defined by \(a(n):=\sum_{d^ k| n}\mu (d)\tau (n/d^ k)\) (k\(\in {\mathbb{N}}\), arithmetical function \(\tau\) (n)\(\geq 0)\). The generating function Z(s) of \(\tau\) ( ) has to satisfy certain growth restrictions in an appropriate halfplane of the type `Lindelöf hypothesis'. Finally this result is applied to obtain `nice' asymptotic formulae for the summatory functions of the (primitive) Pythagorean triangles.