\documentclass[11pt,a4paper]{article} \usepackage[utf8]{inputenc} \usepackage{amsmath,amssymb,amsthm} \usepackage{geometry} \usepackage{booktabs} \usepackage{hyperref} \usepackage{tikz} \usepackage{graphicx} \usepackage{float} \usepackage{caption} \geometry{margin=0.95in} \hypersetup{colorlinks=true, linkcolor=blue, citecolor=blue, urlcolor=blue} \newtheorem{theorem}{Theorem}[section] \newtheorem{proposition}[theorem]{Proposition} \newtheorem{corollary}[theorem]{Corollary} \renewcommand{\labelitemi}{$\bullet$} \title{\textbf{Fourier Epicycles and Antisymmetric Signed-Area Conservation in Normalized Riemann Zeta Terms}} \author{Ryan Hackbarth \\ \texttt{\href{mailto:hackbarthr16@hanover.edu}{hackbarthr16@hanover.edu}}} \date{November 19, 2026} \begin{document} \maketitle \begin{abstract} In companion studies of Euler's alternating series $\eta(s) = (1 - 2^{1-s})\zeta(s)$ and its prime-base normalizations $\eta_p(s) = (1 - p^{1-s})\zeta(s)$, distributing the prefactor $(1 - p^{1-s})^{-1}$ across individual summands $t_{p,n}(s) = c_p(n)(1 - p^{1-s})^{-1}n^{-s}$ exposed an infinite hierarchy of closed, irreducible degree-$2k$ plane algebraic curves $t_{p,p^k}(\frac{1}{2}+ib)$ over $\mathbb{Z}[x, y]$ on the critical line $\operatorname{Re}(s) = \frac{1}{2}$, whose enclosed signed areas satisfy $\mathcal{A}_{\mathrm{signed}}(t_4) = 0$ at the base-$2$ Bernoulli lemniscate and $\sum_{k=1}^{\infty}\mathcal{A}_{\mathrm{signed}}(t_{p,p^k}) = 0$ across every prime tower. Rather than keeping $(1 - p^{1-s})^{-1}$ unexpanded to preserve Cartesian curve equations, this paper expands $(1 - p^{1-s})^{-1} = -\sum_{m=1}^{\infty} p^{m(s-1)}$ across the critical strip $0 < \operatorname{Re}(s) < 1$ and distributes it termwise as a Fourier epicycle series. For every pure power $n = p^k$ ($k \ge 1$) along $s = a + ib$, the distributed expansion separates into $k - 1$ clockwise circles ($1 \le m \le k - 1$), a motionless center-of-mass anchor $+(p - 1)p^{-k}$ ($m = k$) independent of $s$, and an infinite counterclockwise tail ($m \ge k + 1$), revealing index-by-index how counter-rotating harmonics fold and area-balance the Lemniscate of Bernoulli $t_4(\frac{1}{2}+ib)$. Evaluating Green's theorem via Parseval's identity yields an exact closed-form signed-area formula for $t_{p,p^k}(a+ib)$ across $0 < a < 1$, proving that $t_4(a+ib)$ encloses zero net signed area if and only if $a = \frac{1}{2}$---tipping toward its clockwise left lobe for $0 < a < \frac{1}{2}$ and its counterclockwise right lobe for $\frac{1}{2} < a < 1$. Summing over $k \ge 1$ further proves that the total pure-power signed area $\mathcal{S}_p(a)$ is strictly antisymmetric across the critical line ($\mathcal{S}_p(1 - a) = -\mathcal{S}_p(a)$), negative on $0 < a < \frac{1}{2}$, positive on $\frac{1}{2} < a < 1$, and zero if and only if $\operatorname{Re}(s) = \frac{1}{2}$. \end{abstract} \section{Introduction and Critical-Line Launch Point} The Dirichlet eta function $\eta(s) = (1 - 2^{1-s})\zeta(s)$ provides the classical analytic continuation of the Riemann zeta function $\zeta(s)$ across the critical strip $0 < \operatorname{Re}(s) < 1$, and extends naturally to any prime base $p$ through the balanced Dirichlet series \cite{Apostol1976,Edwards1974,Euler1749,Hackbarth2026b,Titchmarsh1986}: \begin{equation} \label{eq:eta_p_def} \zeta(s) = \frac{\eta_p(s)}{1 - p^{1-s}}, \qquad \eta_p(s) = \sum_{n=1}^{\infty} \frac{c_p(n)}{n^s}, \qquad c_p(n) = \begin{cases} 1, & p \nmid n, \\ -(p - 1), & p \mid n. \end{cases} \end{equation} In two companion studies \cite{Hackbarth2026a,Hackbarth2026b}, distributing the normalizing prefactor $(1 - p^{1-s})^{-1}$ directly into each summand defined the normalized term sequence (with $t_n(s) := t_{2,n}(s)$ in base $2$): \begin{equation} \label{eq:tpn_def} t_{p,n}(s) = \frac{c_p(n)}{(1 - p^{1-s})n^s}, \qquad s = a + ib \quad (a \in (0, 1)). \end{equation} Keeping the denominator $(1 - p^{1-s})$ intact inside Equation \eqref{eq:tpn_def} exposed a sharp geometric division on the critical line $\operatorname{Re}(s) = \frac{1}{2}$. Weighting non-power indices ($n \neq p^k$) by $(1 - p^{-s})$ collapses precessing trajectories into concentric origin-centered circles of constant radius $|c_p(n)|/\sqrt{pn}$ uniquely at $a = \frac{1}{2}$ \cite{GarciaMashreghiRoss2018,Hackbarth2026a,Hackbarth2026b,Needham1997}. Meanwhile, the pure-power terms $t_{p,p^k}(\frac{1}{2}+ib)$ ($k \ge 1$) trace closed, irreducible plane algebraic curves of degree $2k$ over $\mathbb{Z}[x, y]$ whose rational Cartesian equations single out $a = \frac{1}{2}$ among rational abscissas $a \in (0, 1) \cap \mathbb{Q}$, anchored in base $2$ at $k = 2$ by the Lemniscate of Bernoulli $t_4(\frac{1}{2}+ib)$ \cite{Hackbarth2026a,Hackbarth2026b,Hilton1920,Lawrence1972}. At the close of those critical-line analyses, the pure-power hierarchy exhibited an exact oriented-area conservation law \cite{Hackbarth2026a,Hackbarth2026b}: parameterized by $\theta = b\ln p \in [0, 2\pi]$, each curve $t_{p,p^k}(\frac{1}{2}+ib)$ ($k \ge 1$) encloses the signed area \begin{equation} \label{eq:crit_area_law} \mathcal{A}_{\mathrm{signed}}\!\left(t_{p,p^k}\!\left(\frac{1}{2}+ib\right)\right) = \frac{p - k(p - 1)}{p^k}\,\pi, \qquad \sum_{k=1}^{\infty} \mathcal{A}_{\mathrm{signed}}\!\left(t_{p,p^k}\!\left(\frac{1}{2}+ib\right)\right) = 0. \end{equation} On $\operatorname{Re}(s) = \frac{1}{2}$, Equation \eqref{eq:crit_area_law} vanishes individually among all primes and powers solely at $p = 2, k = 2$ ($\mathcal{A}_{\mathrm{signed}}(t_4) = 0$), while across every prime tower $p \ge 2$ the positive area of the base circle $t_{p,p}$ ($k = 1$) balances the negative areas of all higher inner-looped curves ($k \ge 2$). Equation \eqref{eq:crit_area_law} serves as our launch point here. Because those earlier derivations worked primarily on $\operatorname{Re}(s) = \frac{1}{2}$ while keeping $(1 - p^{1-s})^{-1}$ unexpanded to preserve algebraic curve equations, they left open how the enclosed signed areas behave on off-critical vertical lines $a \neq \frac{1}{2}$. In Section \ref{sec:epicycles}, we expand $(1 - p^{1-s})^{-1}$ as a convergent exterior power series across $0 < \operatorname{Re}(s) < 1$ and distribute it across the summands of $\eta_p(s)$. Although termwise expansion dissolves individual closed algebraic curves into uniform-speed circles, viewing the distributed sum as a Fourier epicycle decomposition separates every pure-power curve $t_{p,p^k}(a+ib)$ into finite clockwise rotors, a motionless real-axis anchor, and an infinite counterclockwise tail, illuminating index-by-index how the Bernoulli lemniscate $t_4(\frac{1}{2}+ib)$ is assembled. In Section \ref{sec:area_law}, we apply Parseval's area identity to derive closed-form strip-wide formulas for $\mathcal{A}_{\mathrm{signed}}(t_{p,p^k}(a+ib))$ and its tower sum $\mathcal{S}_p(a)$, proving that both the lobe balance of $t_4(a+ib)$ and prime-tower signed-area conservation hold if and only if $\operatorname{Re}(s) = \frac{1}{2}$, with strict antisymmetry under $a \leftrightarrow 1 - a$. \section{Exterior Power-Series Distribution and Fourier Epicycle Decomposition} \label{sec:epicycles} \subsection{Convergent Strip Expansion and Termwise Epicycle Splitting} To represent the normalizing prefactor $(1 - p^{1-s})^{-1}$ as a convergent power series on vertical lines $s = a + ib$ inside the critical strip $0 < a < 1$, observe that the naive geometric variable $u = p^{1-s}$ has modulus $|u| = p^{1-a} > 1$ throughout $0 < a < 1$ and therefore lies strictly outside the unit Maclaurin disk $|u| < 1$. Likewise, rational substitutions around a fixed center---such as writing $\eta(s) = \frac{x}{1-x}\zeta(s)$ with $x(s) = \frac{1 - 2^{1-s}}{2 - 2^{1-s}}$ and expanding $\frac{1-x}{x} = \sum_{m=1}^{\infty}(1 - x)^m$---cannot converge uniformly along $a = \frac{1}{2}$: on the critical line, $x(\frac{1}{2}+ib)$ traces the origin-centered circle $|x| = \frac{1}{\sqrt{2}}$ enclosing the pole $x = 0$ of $\frac{1-x}{x}$, forcing $|1 - x(\frac{1}{2}+ib)| = (6 - 4\sqrt{2}\cos(b\ln 2))^{-1/2}$ to exceed $1$ whenever $\cos(b\ln 2) \ge \frac{5\sqrt{2}}{8}$. Inverting $u = p^{1-s}$ into the exterior Laurent variable $y := u^{-1} = p^{s-1}$ moves the pole $y = 1$ outside the trajectory across the entire strip $0 < a < 1$, since $|y| = p^{a-1} < 1$ is independent of the ordinate $b$ (with constant modulus $p^{-1/2} < 1$ everywhere on $a = \frac{1}{2}$). Factoring $-p^{1-s}$ from the denominator yields the uniformly convergent strip power series \begin{equation} \label{eq:prefactor_series} \frac{1}{1 - p^{1-s}} = -\frac{p^{s-1}}{1 - p^{s-1}} = -\sum_{m=1}^{\infty} p^{m(s-1)} \qquad (0 < \operatorname{Re}(s) < 1), \end{equation} expressing $\zeta(s)$ across $0 < \operatorname{Re}(s) < 1$ as the convergent product \begin{equation} \label{eq:zeta_prod_series} \zeta(s) = -\left(\sum_{m=1}^{\infty} p^{m(s-1)}\right)\left(\sum_{n=1}^{\infty} \frac{c_p(n)}{n^s}\right). \end{equation} Because $\eta_p(s)$ converges only conditionally on $0 < \operatorname{Re}(s) < 1$, the outer and inner summations in Equation \eqref{eq:zeta_prod_series} cannot be interchanged globally across all $n \in \mathbb{N}$ at once. Distributing the $m$-series into each individual summand $c_p(n)n^{-s}$, however, converges absolutely and uniformly in $b \in \mathbb{R}$ for every fixed $n \ge 1$ and $a \in (0, 1)$: \begin{equation} \label{eq:distributed_general} t_{p,n}(a + ib) = \sum_{m=1}^{\infty} u_{p;n,m}(a + ib), \qquad u_{p;n,m}(a + ib) := -c_p(n)\,p^{-m(1-a)}n^{-a}e^{ib(m\ln p - \ln n)}. \end{equation} Viewed geometrically as $b$ varies, Equation \eqref{eq:distributed_general} trades single algebraic equations for a superposition of uniform-speed circular rotors: \begin{itemize} \item \textbf{Non-Power Indices ($n \neq p^k$):} Since $p^m \neq n$ for every $m \ge 1$, the angular velocity $m\ln p - \ln n$ never vanishes. Every sub-term $u_{p;n,m}(a+ib)$ traces an origin-centered circle of radius $|c_p(n)|p^{-m(1-a)}n^{-a}$, and the curve's time-averaged center of mass $\lim_{T\to\infty}\frac{1}{2T}\int_{-T}^{T} t_{p,n}(a+ib)\,db$ equals $0$ across all $a \in (0, 1)$. \item \textbf{Pure-Power Indices ($n = p^k$, $k \ge 1$):} Here $-c_p(p^k) = p - 1$ and $m\ln p - \ln(p^k) = (m - k)\ln p$. Setting $\theta = b\ln p \in [0, 2\pi]$ locks all harmonics to integer multiples of a single fundamental frequency, partitioning $t_{p,p^k}(a+ib)$ into three distinct kinematic components: \end{itemize} \begin{equation} \label{eq:three_part_split} \begin{aligned} t_{p,p^k}(a + ib) &= \underbrace{(p - 1)\sum_{m=1}^{k-1} p^{-m(1-a)-ka}e^{-i(k-m)\theta}}_{\text{$k - 1$ clockwise circles ($m < k$)}} \;+\; \underbrace{\frac{p - 1}{p^k}}_{\substack{\text{static anchor} \\ (m = k)}} \\ &\quad +\; \underbrace{(p - 1)\sum_{m=k+1}^{\infty} p^{-m(1-a)-ka}e^{+i(m-k)\theta}}_{\text{infinite counterclockwise tail ($m > k$)}}. \end{aligned} \end{equation} Remarkably, at the resonant diagonal index $m = k$ ($p^m = p^k$), both the ordinate phase $(m - k)b\ln p$ and the abscissa $a$ cancel out of $-c_p(p^k)p^{-k(1-a)-ka} = (p - 1)p^{-k}$. Every pure-power curve $t_{p,p^k}(a+ib)$ therefore possesses an $s$-invariant time-averaged center of mass at $\big(+\frac{p-1}{p^k}, 0\big)$ across the entire strip $0 < a < 1$ (summing over $k \ge 1$ to $+1$). In base $2$, this places the mean position of $t_2(\frac{1}{2}+ib)$ at $(+\frac{1}{2}, 0)$ and of $t_4(\frac{1}{2}+ib)$ at $(+\frac{1}{4}, 0)$, shifted left of their geometric centers $(+1, 0)$ and $(+\frac{1}{2}, 0)$ because non-uniform Blaschke speed lingers near periapsis ($\theta = \pi$) and whips rapidly past apoapsis ($\theta = 0$) \cite{Hackbarth2026a}. \subsection{\texorpdfstring{Index-by-Index Construction of the Bernoulli Lemniscate $t_4(\frac{1}{2}+ib)$}{Index-by-Index Construction of the Bernoulli Lemniscate t4(1/2+ib)}} Evaluating Equation \eqref{eq:three_part_split} at $p = 2$, $k = 2$, and $a = \frac{1}{2}$ ($\theta = b\ln 2$) exposes the internal Fourier epicycle structure of the Lemniscate of Bernoulli $t_4(\frac{1}{2}+ib)$: \begin{equation} \label{eq:t4_epicycles} t_4\!\left(\frac{1}{2} + ib\right) = \underbrace{\frac{1}{2\sqrt{2}}e^{-i\theta}}_{\text{CW ($m=1$)}} \;+\; \underbrace{\frac{1}{4}}_{\substack{\text{anchor} \\ (m=2)}} \;+\; \underbrace{\frac{1}{4\sqrt{2}}e^{+i\theta} + \frac{1}{8}e^{+2i\theta} + \frac{1}{8\sqrt{2}}e^{+3i\theta} + \cdots}_{\text{CCW tail ($m \ge 3$)}}. \end{equation} Adding the harmonics of Equation \eqref{eq:t4_epicycles} sequentially by index $m$ reveals how uniform-speed circles fold into a zero-area figure-eight: \begin{enumerate} \item \textbf{Anchor and Clockwise Rotor ($m = 1, 2$):} Starting from the motionless $m = 2$ anchor at $(+\frac{1}{4}, 0)$, the $m = 1$ harmonic attaches a single clockwise circle ($\omega = -1$) of radius $\frac{1}{2\sqrt{2}} \approx 0.3536$, enclosing negative signed area $-\pi\big(\frac{1}{2\sqrt{2}}\big)^2 = -\frac{\pi}{8}$. \item \textbf{First Counterclockwise Rotor ($+m = 3$):} Adding $\frac{1}{4\sqrt{2}}e^{+i\theta}$ ($\omega = +1$) couples two opposite-spinning circles of equal speed around $(+\frac{1}{4}, 0)$. Their horizontal components reinforce to semi-axis $\frac{3}{4\sqrt{2}} \approx 0.5303$ while their vertical components partially cancel to semi-axis $\frac{1}{4\sqrt{2}} \approx 0.1768$, flattening the circle into a clockwise ellipse of net signed area $-\frac{\pi}{8} + \frac{\pi}{32} = -\frac{3\pi}{32}$. \item \textbf{Second Counterclockwise Rotor ($+m = 4$):} Attaching the double-speed rotor $\frac{1}{8}e^{+2i\theta}$ ($\omega = +2$) pulls the right arc outward at $\theta = 0, \pi$ while pinching the vertical waist across the real axis near $\theta = \pm\frac{\pi}{4}$, producing the first self-crossing figure-eight of net signed area $-\frac{3\pi}{32} + 2\cdot\frac{\pi}{64} = -\frac{\pi}{16}$. \item \textbf{Full Counterclockwise Tail ($m \ge 3$):} Each harmonic $m = 2 + j$ ($j \ge 1$) has radius $2^{-(j+4)/2}$ and winding $+j$, contributing positive signed area $+j\pi\,2^{-(j+4)} = \frac{\pi}{32}\,j\,2^{-(j-1)}$. Although every individual counterclockwise circle is smaller than the $m = 1$ clockwise circle, their faster winding numbers amplify their area contributions so that \begin{equation} \label{eq:t4_tail_balance} \mathcal{A}_{\mathrm{CCW}} = \frac{\pi}{32}\sum_{j=1}^{\infty} j\,2^{-(j-1)} = \frac{\pi}{32}\cdot\frac{1}{(1 - \frac{1}{2})^2} = +\frac{\pi}{8}, \end{equation} canceling $\mathcal{A}_{\mathrm{CW}} = -\frac{\pi}{8}$ identically to close the curve into the symmetric Bernoulli lemniscate spanning $\big[-\frac{\sqrt{2}-1}{2}, \frac{\sqrt{2}+1}{2}\big]$ on the real axis with self-intersection node at $(+\frac{1}{2}, 0)$ \cite{Hackbarth2026a,Lawrence1972}. \end{enumerate} \section{Strip-Wide Signed-Area Law and Critical-Line Antisymmetry} \label{sec:area_law} \subsection{\texorpdfstring{Closed-Form Signed Area of $t_{p,p^k}(a+ib)$ and Lobe Tipping of $t_4(a+ib)$}{Closed-Form Signed Area of tp,pk(a+ib) and Lobe Tipping of t4(a+ib)}} Moving off the critical line ($a \neq \frac{1}{2}$) rescales the clockwise and counterclockwise radii in Equation \eqref{eq:three_part_split} asymmetrically through $p^{-m(1-a)-ka}$, tipping the area balance between opposite-spinning rotors. Applying Hurwitz's and Gronwall's classical Fourier--Green area formula \cite{Gronwall1914,Hurwitz1902} across $0 < a < 1$ yields an exact closed form for every pure-power curve. \begin{theorem}[Strip-Wide Signed-Area Formula and Uniqueness of $t_4$] \label{thm:strip_area} For every integer base $p \ge 2$, exponent $k \ge 1$, and abscissa $a \in (0, 1)$, the closed curve $\theta \mapsto t_{p,p^k}(a + ib)$ ($\theta = b\ln p \in [0, 2\pi]$) encloses the oriented signed area \begin{equation} \label{eq:strip_area_closed} \mathcal{A}_{\mathrm{signed}}\!\big(t_{p,p^k}(a + ib)\big) = \frac{\pi(p - 1)^2\,p^{-2ka}\,p^{2a-2}}{\big(1 - p^{2a-2}\big)^2}\Big(1 - k\big(1 - p^{2a-2}\big)\Big). \end{equation} In particular, for $p = 2$ and $k = 2$, the figure-eight curve $t_4(a + ib)$ crosses itself on the real axis at $x_{\mathrm{node}}(a) = 4^{-a}$ ($\cos\theta = 2^{-a}$) for every $a \in (0, 1)$, and its enclosed signed area \begin{equation} \label{eq:t4_strip_area} \mathcal{A}_{\mathrm{signed}}\!\big(t_4(a + ib)\big) = \frac{\pi\,4^{-a-1}\big(2\cdot 4^{a-1} - 1\big)}{\big(1 - 4^{a-1}\big)^2} \end{equation} is strictly negative for $0 < a < \frac{1}{2}$, strictly positive for $\frac{1}{2} < a < 1$, and vanishes if and only if $a = \frac{1}{2}$. \end{theorem} \begin{proof} For fixed $a \in (0, 1)$, the Fourier series $t_{p,p^k}(a + ib) = \sum_{m=1}^{\infty} c_{k,m}(a)e^{i(m-k)\theta}$ in Equation \eqref{eq:three_part_split} has coefficients $c_{k,m}(a) = (p - 1)p^{-m(1-a)-ka}$, which decay exponentially as $O(p^{-m(1-a)})$ since $1 - a > 0$. Termwise differentiation with respect to $\theta$ is therefore uniformly valid, and Green's theorem combined with $L^2([0, 2\pi])$ orthogonality of $\{e^{in\theta}\}_{n\in\mathbb{Z}}$ \cite{Gronwall1914,Hurwitz1902} gives \begin{equation} \label{eq:parseval_step} \begin{aligned} \mathcal{A}_{\mathrm{signed}}\!\big(t_{p,p^k}(a + ib)\big) &= \frac{1}{2}\int_0^{2\pi} \operatorname{Im}\!\left(\overline{t_{p,p^k}(a+ib)}\,\frac{\partial t_{p,p^k}(a+ib)}{\partial\theta}\right) d\theta \\ &= \pi\sum_{m=1}^{\infty} (m - k)\,\big|c_{k,m}(a)\big|^2 = \pi(p - 1)^2 p^{-2ka} \sum_{m=1}^{\infty} (m - k)\,q^m, \end{aligned} \end{equation} where $q := p^{-2(1-a)} = p^{2a-2} \in (p^{-2}, 1)$ for $a \in (0, 1)$. Substituting the convergent geometric sums $\sum_{m=1}^{\infty} m\,q^m = \frac{q}{(1-q)^2}$ and $\sum_{m=1}^{\infty} q^m = \frac{q}{1-q}$ into Equation \eqref{eq:parseval_step} yields \begin{equation*} \pi(p - 1)^2 p^{-2ka}\left(\frac{q}{(1 - q)^2} - \frac{k\,q}{1 - q}\right) = \frac{\pi(p - 1)^2 p^{-2ka}\,q}{(1 - q)^2}\big(1 - k(1 - q)\big), \end{equation*} establishing Equation \eqref{eq:strip_area_closed} (which reduces at $a = \frac{1}{2}$, $q = p^{-1}$, to $\frac{p - k(p-1)}{p^k}\pi$ in Equation \eqref{eq:crit_area_law}). Setting $p = 2$, $k = 2$, and $z = e^{-i\theta}$ in $t_4(a+ib) = -4^{-a}z^2(1 - 2^{1-a}z)^{-1}$, multiplying numerator and denominator by $1 - 2^{1-a}\bar{z}$ gives \begin{equation*} \operatorname{Im} t_4(a + ib) = \frac{4^{-a}\sin\theta\,\big(2\cos\theta - 2^{1-a}\big)}{1 + 4^{1-a} - 2^{2-a}\cos\theta}. \end{equation*} Since $2^{1-a}/2 = 2^{-a} \in (\frac{1}{2}, 1)$ for all $a \in (0, 1)$, $\operatorname{Im} t_4(a+ib) = 0$ occurs at the periapsis $\theta = \pi$, the apoapsis $\theta = 0$, and the non-tangential self-intersection pair $\theta = \pm\arccos(2^{-a})$, where \begin{equation*} t_4\big(a + i\theta/\ln 2\big)\Big|_{\cos\theta = 2^{-a}} = \frac{-4^{-a}\big(\cos 2\theta - 2^{1-a}\cos\theta\big)}{1 + 4^{1-a} - 2^{2-a}(2^{-a})} = \frac{-4^{-a}\big(2\cdot 4^{-a} - 1 - 2\cdot 4^{-a}\big)}{1} = 4^{-a}. \end{equation*} Evaluating Equation \eqref{eq:strip_area_closed} at $p = 2, k = 2$ ($q = 4^{a-1}$) gives Equation \eqref{eq:t4_strip_area}, whose sign is governed by $2\cdot 4^{a-1} - 1 = 2^{2a-1} - 1$, which is negative on $(0, \frac{1}{2})$, positive on $(\frac{1}{2}, 1)$, and zero uniquely at $a = \frac{1}{2}$. \end{proof} Figure \ref{fig:t4_lobes} illustrates Theorem \ref{thm:strip_area} across $a \in \{0.25, 0.50, 0.75\}$. While the $m = 2$ Fourier anchor stays fixed at $(+\frac{1}{4}, 0)$ across all three panels, increasing $a$ slides the self-intersection node $x_{\mathrm{node}}(a) = 4^{-a}$ leftward from $+0.707$ ($a = 0.25$, clockwise left lobe dominant) through $+0.500$ ($a = \frac{1}{2}$, congruent Bernoulli lemniscate lobes) to $+0.354$ ($a = 0.75$, counterclockwise right lobe dominant). \begin{figure}[H] \centering \includegraphics[width=\textwidth]{fig1_t4_lobe_signed_area.png} \caption{Off-critical lobe tipping and enclosed signed area of $t_4(a + ib)$ across $0 < a < 1$: (a) at $a = 0.25 < \frac{1}{2}$, the self-intersection node sits at $(4^{-0.25}, 0) \approx (+0.707, 0)$ and the clockwise left lobe dominates ($\mathcal{A}_{\mathrm{signed}} \approx -0.3892$); (b) on the critical line $a = \frac{1}{2}$, the node lands at $(+0.500, 0)$, balancing both lobes into a Lemniscate of Bernoulli ($\mathcal{A}_{\mathrm{signed}} = 0$) while the $m = 2$ Fourier anchor remains at $(+\frac{1}{4}, 0)$; and (c) at $a = 0.75 > \frac{1}{2}$, the node retreats to $(4^{-0.75}, 0) \approx (+0.354, 0)$ and the counterclockwise right lobe dominates ($\mathcal{A}_{\mathrm{signed}} \approx +1.3408$).} \label{fig:t4_lobes} \end{figure} \subsection{Antisymmetric Zero-Crossing Across Prime Towers} For odd primes $p \ge 3$, $1 - k(1 - p^{-1}) = 0$ solves to $k = \frac{p}{p-1} \notin \mathbb{Z}$, so no single curve $t_{p,p^k}(\frac{1}{2}+ib)$ encloses zero area on the critical line \cite{Hackbarth2026b}. Summing Theorem \ref{thm:strip_area} across the full pure-power tower $k \ge 1$, however, extends the critical-line conservation law \eqref{eq:crit_area_law} to an antisymmetric characterization of $\operatorname{Re}(s) = \frac{1}{2}$ valid for every base $p \ge 2$ (Figure \ref{fig:tower_antisymmetry}). \begin{corollary}[Antisymmetric Prime-Tower Signed-Area Law] \label{cor:tower_antisym} For every integer base $p \ge 2$ and vertical-line abscissa $a \in (0, 1)$, the total signed area of the pure-power curve tower $\mathcal{S}_p(a) := \sum_{k=1}^{\infty} \mathcal{A}_{\mathrm{signed}}\big(t_{p,p^k}(a + ib)\big)$ converges absolutely to \begin{equation} \label{eq:tower_sum_closed} \mathcal{S}_p(a) = \frac{\pi(p - 1)^2\big(p^{2a-2} - p^{-2a}\big)}{p^2\big(1 - p^{2a-2}\big)^2\big(1 - p^{-2a}\big)^2}. \end{equation} Consequently, $\mathcal{S}_p(a)$ satisfies the reflection antisymmetry $\mathcal{S}_p(1 - a) = -\mathcal{S}_p(a)$ for all $a \in (0, 1)$, with $\mathcal{S}_p(a) < 0$ for $0 < a < \frac{1}{2}$, $\mathcal{S}_p(a) > 0$ for $\frac{1}{2} < a < 1$, and $\mathcal{S}_p(a) = 0$ if and only if $a = \frac{1}{2}$. \end{corollary} \begin{proof} Let $r := p^{-2a} \in (p^{-2}, 1)$ and $q := p^{2a-2} \in (p^{-2}, 1)$, noting that $qr = p^{-2}$. Since $\big|\mathcal{A}_{\mathrm{signed}}(t_{p,p^k}(a+ib))\big| = O(k\,r^k)$ and $0 < r < 1$, summing Equation \eqref{eq:strip_area_closed} over $k \ge 1$ via $\sum_{k=1}^{\infty} r^k = \frac{r}{1-r}$ and $\sum_{k=1}^{\infty} k\,r^k = \frac{r}{(1-r)^2}$ gives \begin{equation*} \begin{aligned} \mathcal{S}_p(a) &= \frac{\pi(p - 1)^2 q}{(1 - q)^2}\left(\frac{r}{1 - r} - \frac{(1 - q)r}{(1 - r)^2}\right) \\ &= \frac{\pi(p - 1)^2 qr}{(1 - q)^2(1 - r)^2}\big((1 - r) - (1 - q)\big) = \frac{\pi(p - 1)^2(q - r)}{p^2(1 - q)^2(1 - r)^2}, \end{aligned} \end{equation*} proving Equation \eqref{eq:tower_sum_closed}. Under $a \mapsto 1 - a$, $q = p^{2a-2}$ and $r = p^{-2a}$ swap places, reversing the sign of $q - r$ while leaving $p^2(1 - q)^2(1 - r)^2 > 0$ invariant, so $\mathcal{S}_p(1 - a) = -\mathcal{S}_p(a)$. Finally, since $p > 1$, $p^{2a-2} - p^{-2a}$ has the same sign as $(2a - 2) - (-2a) = 4(a - \frac{1}{2})$, which is negative on $(0, \frac{1}{2})$, positive on $(\frac{1}{2}, 1)$, and zero exclusively at $a = \frac{1}{2}$. \end{proof} \begin{figure}[H] \centering \includegraphics[width=0.86\textwidth]{fig2_antisymmetric_tower_area.png} \caption{Total pure-power signed area $\mathcal{S}_p(a) = \sum_{k=1}^{\infty} \mathcal{A}_{\mathrm{signed}}(t_{p,p^k}(a+ib))$ across $a \in [0.20, 0.80]$ for prime bases $p \in \{2, 3, 5, 7\}$. Every prime tower is strictly negative on $0 < a < \frac{1}{2}$ (clockwise epicycles dominant), strictly positive on $\frac{1}{2} < a < 1$ (counterclockwise tail dominant), antisymmetric under $a \leftrightarrow 1 - a$, and crosses zero uniquely on the critical line $a = \frac{1}{2}$.} \label{fig:tower_antisymmetry} \end{figure} Inspecting the double sum behind Corollary \ref{cor:tower_antisym}, \begin{equation} \label{eq:double_sum_symmetry} \mathcal{S}_p(a) = \pi(p - 1)^2 \sum_{k=1}^{\infty}\sum_{m=1}^{\infty} (m - k)\,p^{-2ka - 2m(1-a)}, \end{equation} makes the structural mechanism behind this zero-crossing transparent. The harmonic winding factor $(m - k)$ is antisymmetric under exchanging the Dirichlet index $k$ (governed by $p^{-ks}$) with the exterior prefactor index $m$ (governed by $p^{m(s-1)}$), whereas the squared-modulus weight $p^{-2ka - 2m(1-a)}$ becomes symmetric in $(k, m)$ if and only if $a = 1 - a$, i.e., $a = \frac{1}{2}$. \section{Conclusion} Comparing the unexpanded and power-series representations of $(1 - p^{1-s})^{-1}$ highlights how two elementary faces of the same normalization expose complementary geometric structures inside $\zeta(s)$ across $0 < \operatorname{Re}(s) < 1$. Keeping $(1 - p^{1-s})^{-1}$ intact preserves the closed degree-$2k$ algebraic curves $t_{p,p^k}(\frac{1}{2}+ib)$ over $\mathbb{Z}[x, y]$, their Blaschke elliptic perimeters, and the prime-versus-composite dichotomy established in \cite{Hackbarth2026a,Hackbarth2026b}. Expanding $(1 - p^{1-s})^{-1} = -\sum_{m=1}^{\infty} p^{m(s-1)}$ and distributing it termwise instead resolves each pure-power trajectory into $k - 1$ clockwise rotors, a motionless real-axis center-of-mass anchor $+(p - 1)p^{-k}$, and an infinite counterclockwise tail. Through Green's theorem and Parseval's identity, that counter-rotating decomposition explains both how the Bernoulli lemniscate $t_4(\frac{1}{2}+ib)$ is pinched and area-balanced index-by-index and why $\operatorname{Re}(s) = \frac{1}{2}$ is the unique vertical line in the critical strip where $t_4(a+ib)$ and every prime pure-power tower achieve exact signed-area conservation. \begin{thebibliography}{12} \bibitem{Apostol1976} T.~M. Apostol, \emph{Introduction to Analytic Number Theory}, Undergraduate Texts in Mathematics, Springer-Verlag, New York, 1976. \bibitem{Edwards1974} H.~M. Edwards, \emph{Riemann's Zeta Function}, Academic Press, New York, 1974. \bibitem{Euler1749} L.~Euler, ``Remarques sur un beau rapport entre les s\'eries des puissances tant directes que r\'eciproques,'' \emph{M\'emoires de l'Acad\'emie des Sciences de Berlin}, vol.~17, pp.~83--106, 1768 (read 1749); reprinted in \emph{Opera Omnia}, Series~1, vol.~15, pp.~70--90. \bibitem{GarciaMashreghiRoss2018} S.~R. Garcia, J.~Mashreghi, and W.~T. Ross, \emph{Finite Blaschke Products and Their Connections}, Springer Monographs in Mathematics, Springer, Cham, 2018. \bibitem{Gronwall1914} T.~H. Gronwall, ``Some remarks on conformal representation,'' \emph{Annals of Mathematics}, Second Series, vol.~16, no.~1/4, pp.~72--76, 1914. \bibitem{Hackbarth2026a} R.~Hackbarth, ``Characterizing the Critical Line via Normalized Dirichlet Eta Terms: Concentric Circles, a Bernoulli Lemniscate, and Integer Algebraic Curves,'' preprint, 2026, \url{https://zenodo.org/records/24325291}. \bibitem{Hackbarth2026b} R.~Hackbarth, ``Plane Algebraic Curve Families from Normalized Zeta Series: Prime-Base Extensions and Composite Breakdown,'' preprint, 2026, \url{https://zenodo.org/records/24332600}. \bibitem{Hilton1920} H.~Hilton, \emph{Plane Algebraic Curves}, Clarendon Press, Oxford, 1920. \bibitem{Hurwitz1902} A.~Hurwitz, ``Sur quelques applications g\'eom\'etriques des s\'eries de Fourier,'' \emph{Annales scientifiques de l'\'Ecole Normale Sup\'erieure}, S\'erie~3, vol.~19, pp.~357--408, 1902. \bibitem{Lawrence1972} J.~D. Lawrence, \emph{A Catalog of Special Plane Curves}, Dover Publications, New York, 1972. \bibitem{Needham1997} T.~Needham, \emph{Visual Complex Analysis}, Clarendon Press, Oxford, 1997. \bibitem{Titchmarsh1986} E.~C. Titchmarsh, \emph{The Theory of the Riemann Zeta-Function}, 2nd ed., edited and with a preface by D.~R. Heath-Brown, Clarendon Press, Oxford, 1986. \end{thebibliography} \end{document}