Maslov Anomaly and the Morse Index Theorem

Our starting point is again the phase space integral $$\displaystyle{ \text{e}^{\text{i}\hat{\varGamma }[\tilde{M}]} =\int \mathcal{D}\chi ^{a}\,\text{e}^{\text{i}S_{\text{fl}}[\chi,\tilde{M}]} }$$ (31.1) with periodic boundary conditions χ(0) = χ(T) and $$\displaystyle{ S_{\text{fl}}[\chi,\tilde{M}] = \frac{1} {2}\int _{0}^{T}dt\,\bar{\chi }_{ a}(t)\left [ \frac{\partial } {\partial t} -\tilde{M}(t)\right ]_{\phantom{a}b}^{a}\chi ^{b}(t)\;. }$$ (31.2) Here we have indicated that Sfl and \(\hat{\varGamma }\) depend on ηcl a and A i only through \(\tilde{M}_{\phantom{a}b}^{a}\): $$\displaystyle{ \tilde{M}(t)_{\phantom{a}b}^{a} =\omega ^{ac}\partial _{ c}\partial _{b}\mathcal{H}{\bigl (\eta _…

On Riemann’s Ideas on Space and Schwinger’s Treatment of Low-Energy Pion-Nucleon Physics

We begin with a demonstration of how great an influence Riemann’s habilitation essay had on the development of field theory. His ideas about the origin of physical space and the importance of a metric field were clearly outlined as early as 1854, and praised highly by the old C.F. Gauss, who died 1 year later. There is but one formula in Riemann’s article. This formula and its relevance will be explained at the beginning of the present chapter. The basic principle which is omnipresent in Riemann’s entire work is to understand the physical behavior of nature from its smallness. Hence partial differential equations stand at the beginning of any field theory. In our case, it is not the metric …

Time-Independent Canonical Perturbation Theory

First we consider the perturbation calculation only to first order, limiting ourselves to only one degree of freedom. Furthermore, the system is to be conservative, ∂ H∕∂ t = 0, and periodic in both the unperturbed and perturbed case. In addition to periodicity, we shall require the Hamilton–Jacobi equation to be separable for the unperturbed situation. The unperturbed problem H0(J0) which is described by the action-angle variables J0 and w0 will be assumed to be solved. Thus we have, for the unperturbed frequency: $$\displaystyle{ \nu _{0} = \frac{\partial H_{0}} {\partial J_{0}} }$$ (10.1) and $$\displaystyle{ w_{0} =\nu _{0}t +\beta _{0}\;. }$$ (10.2) Then the new Hamiltonian reads, up t…

Propagators for Particles in an External Magnetic Field

In order to describe the propagation of a scalar particle in an external potential, we begin again with the path integral $$ K(r',t';r,0) = \int_{r,(0)}^{r',(t')} {[dr(t)]} \exp \left\{ {\frac{{\text{i}}} {\hbar }S[r(t)]} \right\} $$ (1) with $$ S[r(t)] = \int_0^{t'} {dt} L(r,\dot r). $$

Poincaré Surface of Sections, Mappings

We consider a system with two degrees of freedom, which we describe in four-dimensional phase space. In this (finite) space we define an (oriented) two-dimensional surface. If we then consider the trajectory in phase space, we are interested primarily in its piercing points through this surface. This piercing can occur repeatedly in the same direction. If the motion of the trajectory is determined by the Hamiltonian equations, then the n + 1-th piercing point depends only on the nth. The Hamiltonian thus induces a mapping n → n + 1 in the “Poincare surface of section” (PSS). The mapping transforms points of the PSS into other (or the same) points of the PSS. In the following we shall limit …

Superconvergent Perturbation Theory, KAM Theorem (Introduction)

Here we are dealing with an especially fast converging perturbation series, which is of particular importance for the proof of the KAM theorem (cf. below).

Partition Function for the Harmonic Oscillator

We start by making the following changes from Minkowski real time t = x0 to Euclidean “time” τ = tE:

Introduction to Homotopy Theory

Consider two manifolds X and Y together with a set of continuous maps f, g,... $$ f:X \to Y,x \to f(x) = y;x \in X,y \in Y. $$

Direct Evaluation of Path Integrals

Every time τ n is assigned a point y n . We now connect the individual points with a classical path y(τ). y(τ) is not necessarily the (on-shell trajectory) extremum of the classical action. It can be any path between τ n and τn−1 specified by the classical Lagrangian \(L(y,\dot{y},t).\)

The Hamilton–Jacobi Equation

We already know that canonical transformations are useful for solving mechanical problems. We now want to look for a canonical transformation that transforms the 2N coordinates (q i , p i ) to 2N constant values (Q i , P i ), e.g., to the 2N initial values \((q_{i}^{0},p_{i}^{0})\) at time t = 0. Then the problem would be solved, q = q(q0, p0, t), p = p(q0, p0, t).

Particle in Harmonic E-Field E(t) = Esinω 0 t; Schwinger–Fock Proper-Time Method

Since the Green’s function of a Dirac particle in an external field, which is described by a potential A μ (x), is given by $$\displaystyle{ \left [\gamma \cdot \left (\frac{1} {i} \partial - eA\right ) + m\right ]G(x,x^{{\prime}}\vert A) =\delta (x - x^{{\prime}}) }$$ (37.1) the Green operator G+[A] is defined by $$\displaystyle{ \left (\gamma \Pi + m\right )G_{+} = 1\,,\quad \Pi _{\mu } = p_{\mu } - eA_{\mu } }$$ or $$\displaystyle\begin{array}{rcl} G_{+}& =& \frac{1} {\gamma \Pi + m - i\epsilon }\,,\quad \epsilon > 0 {}\\ & =& \frac{\gamma \Pi - m} {\left (\gamma \Pi \right )^{2} - m^{2} + i\epsilon } = \frac{-\gamma \Pi + m} {m^{2} -\left (\gamma \Pi \right )^{2} - i\epsilon } {}\\ & =&…

The KAM Theorem

This theorem guarantees that, under certain assumptions, in the case of a perturbation \(\varepsilon H_{1}(\boldsymbol{J},\boldsymbol{\theta })\) with small enough ɛ, the iterated series for the generator W(θ i 0, J i ) converges (according to Newton’s procedure) and thus the invariant tori are not destroyed. The KAM theorem is valid for systems with two and more degrees of freedom. However, in the following, we shall deal exclusively with the case of two degrees of freedom.

Topological Phases in Planar Electrodynamics

This section is meant to be an extension of Chap. 31 on the quantal Berry phases. In particular, we are interested in studying the electromagnetic interaction of particles with a nonzero magnetic moment in \(D = 2 + 1\) dimensions and of translational invariant configurations of \((D = 3 + 1)\)-dimensional charged strings with a nonzero magnetic moment per unit length. The whole discussion is based on our article in Physical Review D44, 1132 (1991).

The Action Principles in Mechanics

We begin this chapter with the definition of the action functional as time integral over the Lagrangian \(L(q_{i}(t),\dot{q}_{i}(t);t)\) of a dynamical system: $$\displaystyle{ S\left \{[q_{i}(t)];t_{1},t_{2}\right \} =\int _{ t_{1}}^{t_{2} }dt\,L(q_{i}(t),\dot{q}_{i}(t);t)\;. }$$

The Action Principle in Classical Electrodynamics

The main purpose of this chapter is to consider the formulation of a relativistic point particle in classical electrodynamics from the viewpoint of Lagrangian mechanics. Here, the utility of Schwinger’s action principle is illustrated by employing three different kinds of action to derive the equations of motion and the associated surface terms.

Linear Oscillator with Time-Dependent Frequency

Here is another important example of a path integral calculation, namely the time-dependent oscillator whose Lagrangian is given by $$\displaystyle{ L = \frac{m} {2} \dot{x}^{2} -\frac{m} {2} W(t)x^{2}\;. }$$ (21.1) Since L is quadratic, we again expand around a classical solution so that later on we will be dealing again with the calculation of the following path integral: $$\displaystyle{ \int _{x(t_{i})\,=\,0}^{x(t_{f})\,=\,0}[dx(t)]\text{exp}\left \{ \frac{\text{i}} {\hslash }\,\frac{m} {2} \int _{t_{i}}^{t_{f} }dt\left [\left (\frac{dx} {dt} \right )^{\!2} - W(t)x^{2}\right ]\right \}\;. }$$ (21.2) Using \(x(t_{i}) = 0 = x(t_{f}),\) we can integrate by parts and obtain $$\displaystyle{…

The WKB Approximation

In this chapter we shall develop an important semiclassical method which has come back into favor again, particularly in the last few years, since it permits a continuation into field theory. Here, too, one is interested in nonperturbative methods.

Berry Phase and Parametric Harmonic Oscillator

Our concern in this section is once more with the time-dependent harmonic oscillator with Lagrangian $$\displaystyle{ L = \frac{1} {2}\dot{x}^{2} -\frac{1} {2}\omega ^{2}(t)x^{2}\;. }$$ To present a coherent picture of the whole problem, let us briefly review some of the results of Chap. 21. There we found the propagation function

Functional Derivative Approach

Let us now leave the path integral formalism temporarily and reformulate operatorial quantum mechanics in a way which will make it easy later on to establish the formal connection between operator and path integral formalism. Our objective is to introduce the generating functional into quantum mechanics. Naturally we want to generate transition amplitudes. The problem confronting us is how to transcribe operator quantum mechanics as expressed in Heisenberg’s equation of motion into a theory formulated solely in terms of c-numbers. This can be achieved either by Schwinger’s action principle or with the aid of a generation functional defined as follows:

Computing the Trace

So far we have been interested in the general expression for the WKB-propagation function. Now we turn our attention to the trace of that propagator, since we want to exhibit the energy eigenvalues of a given potential. From earlier discussions we know that the energy levels of a given Hamiltonian are provided by the poles of the Green’s function:

Riemann’s Result and Consequences for Physics and Philosophy

Riemann commented on his main result as follows: “The common character of those manifolds whose curvature is constant may also be expressed thus: that figures may be viewed in them without stretching. For clearly figures could not be arbitrarily shifted and turned around in them if the curvature at each point were not the same in all directions at one point as at another, and consequently the same constructions can be made from it; whence it follows that in aggregates with constant curvature, figures may have any arbitrary position given them. The measure-relations of these manifolds depend only on the value of the curvature, and in relation to the analytic expression it may be remarked tha…

Removal of Resonances

From the perturbative procedure in the last chapter we have learned that in the proximity of resonances of the unperturbed system, resonant denominators appear in the expression for the adiabatic invariants. We now wish to begin to locally remove such resonances by trying, with the help of a canonical transformation, to go to a coordinate system which rotates with the resonant frequency.

Classical Chern–Simons Mechanics

We are interested in a completely integrable Hamiltonian system \((\mathscr{M}_{2N},\omega,H).\) Local coordinates on the 2N-dimensional phase space \(\mathscr{M}_{2N}\) are denoted by η a = (p, q), a = 1, 2, … 2N and the symplectic 2-form ω is given by

The Usefulness of Lie Brackets: From Classical and Quantum Mechanics to Quantum Electrodynamics

We know that in Hamiltonian systems a dynamic function f(q, p) develops in time according to

Jacobi Fields, Conjugate Points

Let us go back to the action principle as realized by Jacobi, i.e., time is eliminated, so we are dealing with the space trajectory of a particle. In particular, we want to investigate the conditions under which a path is a minimum of the action and those under which it is merely an extremum. For illustrative purposes we consider a particle in two-dimensional real space.

Green’s Function of a Spin- 1 2 $$\tfrac {1}{2}$$ Particle in a Constant External Magnetic Field

Our objective here is to find the Green’s function of a spin-\(\tfrac {1}{2}\) particle in an external electromagnetic field. Accordingly we start with the defining equation

Particle in Harmonic E-Field E ( t ) = E sin ω 0 t $$E(t)= E \sin \omega _0 t$$ ; Schwinger–Fock Proper-Time Method

Since the Green’s function of a Dirac particle in an external field, which is described by a potential Aμ(x), is given by

One-Loop Effective Lagrangian in QED

Our main goal in this section is the derivation of an expression for the effective Lagrangian in one-loop approximation. So let’s start with the vacuum persistence amplitude in presence of an external field: $$\displaystyle \langle 0_+\vert 0_-\rangle ^A = e^{ iW^{(1)}[A]} = e^{i \int d^4x\mathcal {L}^{(1)}(x)} $$

The Adiabatic Invariance of the Action Variables

We shall first use an example to explain the concept of adiabatic invariance. Let us consider a “super ball” of mass m, which bounces back and forth between two walls (distance l) with velocity \(\boldsymbol{v}_{0}\). Let gravitation be neglected, and the collisions with the walls be elastic. If F m denotes the average force onto each wall, then we have $$\displaystyle{ F_{m}T = -\int _{\mathrm{coll.\,time}}f\,dt\;. }$$ (9.1) f is the force acting on the ball during one collision, and T is the time between collisions.

Canonical Adiabatic Theory

In the present chapter we are concerned with systems, the change of which—with the exception of a single degree of freedom—should proceed slowly. (Compare the pertinent remarks about \(\varepsilon\) as slow parameter in Chap. 7) Accordingly, the Hamiltonian reads: $$\displaystyle{ H = H_{0}{\bigl (J,\varepsilon p_{i},\varepsilon q_{i};\varepsilon t\bigr )} +\varepsilon H_{1}{\bigl (J,\theta,\varepsilon p_{i},\varepsilon q_{i};\varepsilon t\bigr )}\;. }$$ (12.1) Here, \((J,\theta )\) designates the “fast” action-angle variables for the unperturbed, solved problem \(H_{0}(\varepsilon = 0),\) and the (p i , q i ) represent the remaining “slow” canonical variables, which do not necessarily have…

Fundamental Principles of Quantum Mechanics

There are two alternative methods of quantizing a system: a) quantization via the Feynman Path Integral (equivalent to Schwinger’s Action Principle); b) canonical quantization.

Berry’s Phase

Let a physical system be described by a Hamiltonian with two sets of variables \(\boldsymbol{r}\) and \(\boldsymbol{R}(t):\, H(\boldsymbol{r},\boldsymbol{R}(t)).\) The dynamical degrees of freedom \(\boldsymbol{r}\) (not necessarily space variables) are also called fast variables. The external time dependence is given by the slowly varying parameters \(\boldsymbol{R}(t) =\{ X(t),\,Y (t),\,\ldots,\,Z(t)\}\); consequently, the \(\boldsymbol{R}(t)\) are called slow variables.

Classical Geometric Phases: Foucault and Euler

In the last chapter we saw how a quantum system can give rise to a Berry phase, by studying the adiabatic round trip of its quantum state on a certain parameter space. Rather than considering what happens to states in Hilbert space, we now turn to classical mechanics, where we are concerned instead with the evolution of the system in configuration space. As a first example, we are interested in the geometric phase of an oscillator that is constrained to a plane that is transported over some surface which moves along a certain path in three-dimensional space. Contrary to determining the Berry phase, there is no adiabatic approximation of the motion along the curve involved. The Foucault phas…

The Non-Abelian Vector Gauge Particle ρ

So far we have shown how pion self-interaction and pion-mass generation are tied together. This phenomenon, a result of the three-dimensional non-linear realization of the triplet isotopic pion field, is a consequence of a dimensional reduction from four-dimensional Euclidean space to a three-dimensional curved isotopic space with positive curvature, i.e., by way of chiral-symmetry breaking and subsequent pionic mass generation in a curved isotopic spin background with K = λ2, λ = 2f0∕mπ, which is taken from low-energy π − N interaction.

Examples for Calculating Path Integrals

We now want to compute the kernel K(b, a) for a few simple Lagrangians. We have already found for the one-dimensional case that $$\displaystyle{ K{\bigl (x_{2},t_{2};x_{1},t_{1}\bigr )} =\int _{ x(t_{1})=x_{1}}^{x(t_{2})=x_{2} }[dx(t)]\,\text{e}^{(\mathrm{i}/\hslash )S} }$$ (19.1) with $$\displaystyle{ S =\int _{ t_{1}}^{t_{2} }dt\,L(x,\dot{x};t)\;. }$$ First we consider a free particle, $$\displaystyle{ L = m\dot{x}^{2}/2\;, }$$ (19.2) and represent an arbitrary path in the form, $$\displaystyle{ x(t) =\bar{ x}(t) + y(t)\;. }$$ (19.3) Here, \(\bar{x}(t)\) is the actual classical path, i.e., solution to the Euler–Lagrange equation: $$\displaystyle{ \frac{\partial L} {\partial x}\Big\vert _{…

Erratum to: Classical and Quantum Dynamics: From Classical Paths to Path Integrals

Canonical Perturbation Theory with Several Degrees of Freedom

We extend the perturbation theory of the previous chapter by going one order further and permitting several degrees of freedom. So let the unperturbed problem H0(J k 0) be solved.

Path Integral Formulation of Quantum Electrodynamics

Let us consider a pure Abelian gauge theory given by the Lagrangian $$\displaystyle\begin{array}{rcl} \mathcal{L}_{\text{photon}}& =& -\frac{1} {4}F_{\mu \nu }F^{\mu \nu } \\ & =& -\frac{1} {4}\left (\partial _{\mu }A_{\nu } - \partial _{\nu }A_{\mu }\right )\left (\partial ^{\mu }A^{\nu } - \partial ^{\nu }A^{\mu }\right ){}\end{array}$$ (36.1) or, after integration by parts, $$\displaystyle\begin{array}{rcl} \mathcal{L}_{\text{photon}}& =& -\frac{1} {2}\left [-\left (\partial _{\mu }\partial ^{\mu }A_{\nu }\right )A^{\nu } + \left (\partial ^{\mu }\partial ^{\nu }A_{\mu }\right )A_{\nu }\right ] \\ & =& \frac{1} {2}A_{\mu }\left [g^{\mu \nu }\square - \partial ^{\mu }\partial ^{\nu }\righ…

The “Maslov Anomaly” for the Harmonic Oscillator

Specializing the discussion of the previous section to the harmonic oscillator we have for \(N = 1,\ \eta ^{a} = (p,x),\ a = 1,2,\ \eta ^{1} \equiv p,\ \eta ^{2} \equiv x\) $$\displaystyle{ H(p,x) = \frac{1} {2}\eta ^{a}\eta ^{a} = \frac{1} {2}{\bigl (p^{2} + x^{2}\bigr )}\;. }$$ (30.1) The only conserved quantity is J = H. In the action we need the combination $$\displaystyle{ \frac{1} {2}\eta ^{a}\omega _{ ab}\dot{\eta }^{b} -\mathcal{H}(\eta ) = \frac{1} {2}\eta ^{a}\left [\omega _{ ab} \frac{d} {dt} -{\bigl ( 1 + A(t)\bigr )}\mathrm{1l}_{ab}\right ]\eta ^{b} }$$ (30.2) and $$\displaystyle{ \tilde{M}_{\phantom{a}b}^{a} =\omega ^{ac}\partial _{ c}\partial _{b}(H + AJ\,) ={\bigl ( 1 + A(t)…

Action-Angle Variables

In the following we will assume that the Hamiltonian does not depend explicitly on time; ∂H/∂t = 0. Then we know that the characteristic function W(q i , P i ) is the generator of a canonical transformation to new constant momenta P i , (all Q i , are ignorable), and the new Hamiltonian depends only on the P i ,: H = K = K(P i ). Besides, the following canonical equations are valid: $$ \dot Q_i = \frac{{\partial K}} {{\partial P_i }} = v_i = const. $$ (1) $$ \dot P_i = \frac{{\partial K}} {{\partial Q_i }} = 0. $$ (2)