<?xml version="1.0" encoding="utf-8"?><?xml-stylesheet type="text/xsl" href="math.xslt"?><html xmlns="http://www.w3.org/1999/xhtml"><head><link rel="stylesheet" type="text/css" href="style.css"/><title>Non-Noether symmetries in Hamiltonian Dynamical Systems</title></head><body><h1>Non-Noether symmetries in Hamiltonian Dynamical Systems</h1><div class="author">George Chavchanidze</div><div class="affiliation">Department of Theoretical Physics,A. Razmadze Institute of Mathematics,1 Aleksidze Street, Tbilisi 0193, Georgia</div><div class="abstract"><strong class="cap">abstract. </strong>We discuss geometric properties of non-Noether symmetries andtheir possible applications in integrable Hamiltonian systems.Correspondence between non-Noether symmetries and conservation lawsis revisited. It is shown that in regular Hamiltonian systemssuch symmetries canonically lead to Lax pairs on the algebraof linear operators on cotangent bundle over the phase space.Relationship between non-Noether symmetries and other widespread geometricmethods of generating conservation laws such as bi-Hamiltonian formalism,bidifferential calculi and Frölicher-Nijenhuis geometry is considered.It is proved that the integrals of motion associated with acontinuous non-Noether symmetry are in involution whenever thegenerator of the symmetry satisfies a certain Yang-Baxter type equation.Action of one-parameter group of symmetry on algebra of integrals of motionis studied and involutivity of group orbits is discussed.Hidden non-Noether symmetries of Toda chain, Korteweg-de Vries equation,Benney system, nonlinear water wave equations and Broer-Kaup systemare revealed and discussed.</div><div class="keywords"><strong class="cap">keywords. </strong>Non-Noether symmetry; Conservation law; bi-Hamiltonian system; Bidifferential calculus; Lax pair; Frölicher-Nijenhuisoperator; Korteweg-de Vries equation; Broer-Kaup system; Benney system; Toda chain</div><div class="msc"><strong class="cap">msc. </strong> 70H33; 70H06; 58J70; 53Z05; 35A30</div><h2>Introduction</h2><div class="paragraph">Symmetries play essential role in dynamical systems, because they usually simplifyanalysis of evolution equations and often provide quite elegant solution of problems that otherwise wouldbe difficult to handle. In Lagrangian and Hamiltonian dynamical systems special role is playedby Noether symmetries — an important class of symmetries that leave action invariantand have some exceptional features. In particular, Noether symmetries deservedspecial attention due to celebrated Noether's theorem, that established correspondencebetween symmetries, that leave action functional invariant, and conservation lawsof Euler-Lagrange equations. This correspondence can be extended to Hamiltoniansystems where it becomes more tight and evident then in Lagrangian case and gives riseto Lie algebra homomorphism between Lie algebra of Noether symmetries and algebra ofconservation laws (that form Lie algebra under Poisson bracket).</div><div class="paragraph">Role of symmetries that are not of Noether type has been suppressed for quite a long time.However, after some publications of Hojman, Harleston, Lutzky and others(see <a href="#r16">[16]</a>, <a href="#r36">[36]</a>, <a href="#r39">[39]</a>,<a href="#r40">[40]</a>, <a href="#r49">[49]</a>-<a href="#r57">[57]</a>)it became clear that non-Noether symmetries also can play important role inLagrangian and Hamiltonian dynamics. In particular, according to Lutzky<a href="#r51">[51]</a>, in Lagrangian dynamics there is definite  correspondence between non-Noether symmetries andconservation laws. Moreover, each generator of non-Noether symmetrymay produce whole family of conservation laws (maximal number of conservation laws that canbe associated with non-Noether symmetry via Lutzky's theorem is equal to the dimension ofconfiguration space of Lagrangian system). This fact makes non-Noether symmetries especiallyvaluable in infinite dimensional dynamical systems, where potentially one can recoverinfinite sequence of conservation laws knowing single generator of non-Noether symmetry.</div><div class="paragraph">Existence of correspondence between non-Noether symmetries and conserved quantitiesraised many questions concerning relationship among this type of symmetries andother geometric structures emerging in theory of integrable models.In particular one could notice suspicious similarity between the method of constructingconservation laws from generator of non-Noether symmetry andthe way conserved quantities are produced in either Lax theory, bi-Hamiltonian formalism,bicomplex approach or Lenard scheme.It also raised natural question whether set of conservation laws associated with non-Noethersymmetry is involutive or not, and since it appeared that in general it may not be involutive,there emerged the need of involutivity criteria, similar to Yang-Baxter equation used in Lax theoryor compatibility condition in bi-Hamiltonian formalism and bicomplex approach.It was also unclear how to construct conservation laws in case of infinite dimensionaldynamical systems where volume forms used in Lutzky's construction are no longer well defined.Some of these questions were addressed in papers <a href="#r11">[11]</a>-<a href="#r14">[14]</a>,while in the present review we would like to summarize all these issues and to provide someexamples of integrable models that possess non-Noether symmetries.</div><div class="paragraph">Review is organized as follows. In first section we briefly recall some aspects of geometricformulation of Hamiltonian dynamics. Further, in second section,  correspondencebetween non-Noether symmetries and integrals of motion in regular Hamiltonian systems isdiscussed. Lutzky's theorem is reformulated in terms of bivector fieldsand alternative derivation of conserved quantities suitable for computations in infinitedimensional Hamiltonian dynamical systems is suggested. Non-Noether symmetries oftwo and three particle Toda chains are used to illustrate general theory.In the subsequent section geometric formulation of Hojman's theorem <a href="#r36">[36]</a>is revisited and some examples are provided. Section 4 reveals correspondence betweennon-Noether symmetries and Lax pairs. It is shown that non-Noether symmetry canonicallygives rise to a Lax pair of certain type. Lax pair is explicitly constructed in termsof Poisson bivector field and generator of symmetry. Examples of Toda chains are discussed.Next section deals with integrability issues. An analogue of Yang-Baxter equationthat, being satisfied by generator of symmetry, ensures involutivity of setof conservation laws produced by this symmetry, is introduced.Relationship between non-Noether symmetries and bi-Hamiltonian systemsis considered in section 6. It is proved that under certain conditions,non-Noether symmetry endows phase space of regular Hamiltonian system withbi-Hamiltonian structure. We also discuss conditions under which non-Noethersymmetry can be "recovered" from bi-Hamiltonian structure.Theory is illustrated by example of Toda chains. Next section is devoted tobicomplexes and their relationship with non-Noether symmetries. Special kindof deformation of De Rham complex induced by symmetry is constructed in terms ofPoisson bivector field and generator of symmetry.Samples of two and three particle Toda chain are discussed.Section 8 deals with Frölicher-Nijenhuis recursion operators.It is shown that under certain condition non-Noether symmetrygives rise to invariant Frölicher-Nijenhuis operator on tangentbundle over phase space.The last section of theoretical part contains some remarks on action of one-parametergroup of symmetry on algebra of integrals of motion. Special attention is devoted toinvolutivity of group orbits.</div><div class="paragraph">Subsequent sections of present review provide examples of integrable modelsthat possess interesting non-Noether symmetries. In particular section 10 revealsnon-Noether symmetry of <math xmlns="http://xml-maiden.com">n</math>-particle Toda chain. Bi-Hamiltonian structure,conservation laws, bicomplex, Lax pair and Frölicher-Nijenhuis recursionoperator of Toda hierarchy are constructed using this symmetry.Further we focus on infinite dimensional integrable Hamiltonian systems emergingin mathematical physics. In section 11 case of Korteweg-de Vriesequation is discussed. Symmetry of this equation is identified and used in constructionof infinite sequence of conservation laws and bi-Hamiltonian structure ofKdV hierarchy.  Next sectionis devoted to non-Noether symmetries of integrable systems of nonlinear water wave equations,such as dispersive water wave system, Broer-Kaup system and dispersiveless long wave system.Last section focuses on Benney system and its non-Noether symmetry, that appears to be local,gives rise to infinite sequence of conserved densities of Benney hierarchy and endows it withbi-Hamiltonian structure.</div><h2>Regular Hamiltonian systems</h2><div class="paragraph">The basic concept in geometric formulation of Hamiltonian dynamicsis notion of symplectic manifold. Such a manifold plays the role ofthe phase space of the dynamical system and therefore many propertiesof the dynamical system can be quite effectively investigated in the frameworkof symplectic geometry. Before we consider symmetries of the Hamiltonian dynamicalsystems, let us briefly recall some basic notions from symplectic geometry.</div><div class="paragraph">The symplectic manifold is a pair <math xmlns="http://xml-maiden.com">(M, ω)</math>where <math xmlns="http://xml-maiden.com">M</math> is smooth even dimensional manifold and <math xmlns="http://xml-maiden.com">ω</math>is a closed<formula xml:id="e1" xmlns="http://xml-maiden.com">dω = 0</formula>and nondegenerate 2-form on <math xmlns="http://xml-maiden.com">M</math>. Being nondegenerate means thatcontraction of arbitrary non-zero vector field with <math xmlns="http://xml-maiden.com">ω</math> does not vanish<formula xml:id="e2" xmlns="http://xml-maiden.com">i<sub>X</sub>ω = 0  ⇔  X = 0</formula>(here <math xmlns="http://xml-maiden.com">i<sub>X</sub></math> denotes contraction of the vector field <math xmlns="http://xml-maiden.com">X</math>with differential form). Otherwise one can say that <math xmlns="http://xml-maiden.com">ω</math>is nondegenerate if its n-th outer power does not vanish(<math xmlns="http://xml-maiden.com">ω<sup>n</sup> ≠ 0</math>) anywhere on <math xmlns="http://xml-maiden.com">M</math>.In Hamiltonian dynamics <math xmlns="http://xml-maiden.com">M</math> is usually phase space of classical dynamical systemwith finite numbers of degrees of freedom and the symplectic form <math xmlns="http://xml-maiden.com">ω</math>is basic object that defines Poisson bracket structure, algebra of Hamiltonian vector fieldsand the form of Hamilton's equations.</div><div class="paragraph">The symplectic form <math xmlns="http://xml-maiden.com">ω</math> naturally defines isomorphism between vector fieldsand differential 1-forms on <math xmlns="http://xml-maiden.com">M</math> (in other words tangent bundle <math xmlns="http://xml-maiden.com">TM</math>of symplectic manifold can be quite naturally identified withcotangent bundle <math xmlns="http://xml-maiden.com">T<sup>*</sup>M</math>).The isomorphic map <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math> from <math xmlns="http://xml-maiden.com">TM</math> into<math xmlns="http://xml-maiden.com">T<sup>*</sup>M</math> is obtained by taking contractionof the vector field with <math xmlns="http://xml-maiden.com">ω</math><formula xml:id="e3" xmlns="http://xml-maiden.com">Φ<sub>ω</sub>: X  →  − i<sub>X</sub>ω</formula>(minus sign is the matter of convention). This isomorphism gives rise to natural classificationof vector fields. Namely, vector field <math xmlns="http://xml-maiden.com">X<sub>h</sub></math> is said to be Hamiltonianif its image is exact 1-form or in other words if it satisfies Hamilton's equation<formula xml:id="e4" xmlns="http://xml-maiden.com">i<sub>X<sub>h</sub></sub>ω + dh = 0</formula>for some function <math xmlns="http://xml-maiden.com">h</math> on <math xmlns="http://xml-maiden.com">M</math>.Similarly, vector field <math xmlns="http://xml-maiden.com">X</math> is called locally Hamiltonian if it's image is closed 1-form<formula xmlns="http://xml-maiden.com">i<sub>X</sub>ω + u = 0,        du = 0</formula></div><div class="paragraph">One of the nice features of locally Hamiltonian vector fields, known as Liouville's theorem,is that these vector fields preserve symplectic form <math xmlns="http://xml-maiden.com">ω</math>.In other words Lie derivative of the symplectic form <math xmlns="http://xml-maiden.com">ω</math>along arbitrary locally Hamiltonian vector field vanishes<formula xmlns="http://xml-maiden.com">L<sub>X</sub>ω = 0 ⇔ i<sub>X</sub>ω + du = 0,       du = 0</formula>Indeed, using Cartan's formula that expresses Lie derivative in terms of contraction andexterior derivative<formula xmlns="http://xml-maiden.com">L<sub>X</sub> = i<sub>X</sub>d + di<sub>X</sub></formula>one gets<formula xmlns="http://xml-maiden.com">L<sub>X</sub>ω = i<sub>X</sub>dω + di<sub>X</sub>ω =di<sub>X</sub>ω</formula>(since <math xmlns="http://xml-maiden.com">dω = 0</math>) but according to the definition of locally Hamiltonianvector field<formula xmlns="http://xml-maiden.com">di<sub>X</sub>ω = − du = 0</formula>So locally Hamiltonian vector fields preserve <math xmlns="http://xml-maiden.com">ω</math> and vise versa,if vector field preserves symplectic form <math xmlns="http://xml-maiden.com">ω</math> then it is locally Hamiltonian.</div><div class="paragraph">Clearly, Hamiltonian vector fields constitute subset of locally Hamiltonian ones sinceevery exact 1-form is also closed. Moreover one can notice that Hamiltonian vector fields formideal in algebra of locally Hamiltonian vector fields. This fact can be observed as follows.First of all for arbitrary couple of locally Hamiltonian vector fields <math xmlns="http://xml-maiden.com">X, Y</math>we have <math xmlns="http://xml-maiden.com">L<sub>X</sub>ω = L<sub>Y</sub>ω = 0</math> and<formula xmlns="http://xml-maiden.com">L<sub>X</sub>L<sub>Y</sub>ω − L<sub>Y</sub>L<sub>X</sub>ω= L<sub>[X , Y]</sub>ω = 0</formula>so locally Hamiltonian vector fields form Lie algebra (corresponding Lie bracket is ordinarycommutator of vector fields). Further it is clear that for arbitrary Hamiltonian vector field<math xmlns="http://xml-maiden.com">X<sub>h</sub></math> and locally Hamiltonian one <math xmlns="http://xml-maiden.com">Z</math> one has<formula xmlns="http://xml-maiden.com">L<sub>Z</sub>ω = 0</formula>and<formula xmlns="http://xml-maiden.com">i<sub>X<sub>h</sub></sub>ω + dh = 0</formula>that implies<formula xmlns="http://xml-maiden.com">L<sub>Z</sub>(i<sub>X<sub>h</sub></sub>ω + dh) = L<sub>[Z , X<sub>h</sub>]</sub>ω + i<sub>X<sub>h</sub></sub>L<sub>Z</sub>ω +dL<sub>Z</sub>h<line/>= L<sub>[Z , X<sub>h</sub>]</sub>ω + dL<sub>Z</sub>h = 0</formula>thus commutator <math xmlns="http://xml-maiden.com">[Z , X<sub>h</sub>]</math> is Hamiltonian vector field<math xmlns="http://xml-maiden.com">X<sub>L<sub>Z</sub>h</sub></math>,or in other words Hamiltonian vector fields form ideal in algebra of locallyHamiltonian vector fields.</div><div class="paragraph">Isomorphism <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math> can be extended tohigher order vector fields and differential forms by linearity and multiplicativity.Namely,<formula xmlns="http://xml-maiden.com">Φ<sub>ω</sub>(X ∧ Y) =Φ<sub>ω</sub>(X) ∧ Φ<sub>ω</sub>(Y)</formula>Since <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math> is isomorphism, the symplectic form <math xmlns="http://xml-maiden.com">ω</math>has unique counter image <math xmlns="http://xml-maiden.com">W</math> known as Poisson bivector field.Property <math xmlns="http://xml-maiden.com">dω = 0</math> together with non degeneracy implies that bivectorfield <math xmlns="http://xml-maiden.com">W</math> is also nondegenerate (<math xmlns="http://xml-maiden.com">W<sup>n</sup> ≠ 0</math>) and satisfiescondition<formula xml:id="e15" xmlns="http://xml-maiden.com">[W , W] = 0</formula>where bracket <math xmlns="http://xml-maiden.com">[ , ]</math> known as Schouten bracket or supercommutator, is actuallygraded extension of ordinary commutator of vector fields to the case of multivector fields,and can be defined by linearity and derivation property<formula xmlns="http://xml-maiden.com">[C<sub>1</sub> ∧ C<sub>2</sub> ∧ ... ∧ C<sub>n</sub> ,S<sub>1</sub> ∧ S<sub>2</sub> ∧ ... ∧ S<sub>n</sub>] = <line/>(− 1)<sup>p + q</sup>[C<sub>p</sub> , S<sub>q</sub>] ∧C<sub>1</sub> ∧ C<sub>2</sub> ∧ ... ∧ Ĉ<sub>p</sub> ∧ ... ∧ C<sub>n</sub> <line/>∧ S<sub>1</sub> ∧ S<sub>2</sub> ∧ ... ∧ Ŝ<sub>q</sub> ∧ ...∧ S<sub>n</sub></formula>where over hat denotes omission of corresponding vector field.In terms of the bivector field <math xmlns="http://xml-maiden.com">W</math> Liouville's theorem mentioned above can berewritten as follows<formula xml:id="e17" xmlns="http://xml-maiden.com">[W(u) , W] = 0  ⇔  du = 0</formula>for each 1-form <math xmlns="http://xml-maiden.com">u</math>. It follows from graded Jacoby identity satisfied by Schoutenbracket and property <math xmlns="http://xml-maiden.com">[W , W] = 0</math> satisfied by Poisson bivector field.</div><div class="paragraph">Being counter image of symplectic form, <math xmlns="http://xml-maiden.com">W</math> gives rise to map<math xmlns="http://xml-maiden.com">Φ<sub>W</sub></math>, transforming differential 1-forms into vector fields,which is inverted to the map <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math> and is defined by<formula xmlns="http://xml-maiden.com">Φ<sub>W</sub>: u  →  W(u);       Φ<sub>W</sub>Φ<sub>ω</sub> = id</formula>Further we will often use these maps.</div><div class="paragraph">In Hamiltonian dynamical systems Poisson bivector field is geometric object thatunderlies definition of Poisson bracket — kind of Lie bracket on algebra ofsmooth real functions on phase space. In terms of bivector field <math xmlns="http://xml-maiden.com">W</math>Poisson bracket is defined by<formula xml:id="e19" xmlns="http://xml-maiden.com">{f , g} = W(df ∧ dg)</formula>The condition <math xmlns="http://xml-maiden.com">[W , W] = 0</math> satisfied by bivector field ensures thatfor every triple <math xmlns="http://xml-maiden.com">(f, g, h)</math> of smoothfunctions on the phase space the Jacobi identity<formula xml:id="e20" xmlns="http://xml-maiden.com">{f{g , h}} + {h{f , g}} + {g{h , f}} = 0.</formula>is satisfied.Interesting property of the Poisson bracket is that map from algebra of real smooth functionson phase space into algebra of Hamiltonian vector fields, defined by Poisson bivector field<formula xmlns="http://xml-maiden.com">f  →  X<sub>f</sub> = W(df)</formula>appears to be homomorphism of Lie algebras. In other words commutator of two vector fieldsassociated with two arbitrary functions reproduces vector field associated with Poissonbracket of these functions<formula xml:id="e22" xmlns="http://xml-maiden.com">[X<sub>f</sub> , X<sub>g</sub>] = X<sub>{f , g}</sub></formula>This property is consequence of the Liouville theorem and definition of Poisson bracket.Further we also need another useful property of Hamiltonian vector fields and Poisson bracket<formula xml:id="e23" xmlns="http://xml-maiden.com">{f , g} = W(df ∧ dg) = ω(X<sub>f</sub> ∧ X<sub>g</sub>) =L<sub>X<sub>f</sub></sub>g = − L<sub>X<sub>g</sub></sub>g</formula>it also follows from Liouville theoremand definition of Hamiltonian vector fields and Poisson brackets.</div><div class="paragraph">To define dynamics on <math xmlns="http://xml-maiden.com">M</math> one has to specify time evolution of observables(smooth functions on <math xmlns="http://xml-maiden.com">M</math>). In Hamiltonian dynamical systems time evolutionis governed by Hamilton's equation<formula xml:id="e24" xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>f = {h , f}</formula>where <math xmlns="http://xml-maiden.com">h</math> is some fixed smooth function on the phase space called Hamiltonian.In local coordinate frame <math xmlns="http://xml-maiden.com">z<sub>b</sub></math> bivector field <math xmlns="http://xml-maiden.com">W</math>has the form<formula xmlns="http://xml-maiden.com">W = W<sub>bc</sub> <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>c</sub></den></fraction></formula>and the Hamilton's equation rewritten in terms of local coordinates takes the form<formula xmlns="http://xml-maiden.com">ż<sub>b</sub> = W<sub>bc</sub><fraction><num>∂h</num><den>∂z<sub>b</sub></den></fraction></formula></div><h2>Non-Noether symmetries</h2><div class="paragraph">Now let us focus on symmetries of Hamilton's equation <a href="#e24">(24)</a>.Generally speaking, symmetries play very important role in Hamiltonian dynamicsdue to different reasons. They not only give rise to conservation laws butalso often provide very effective solutions to problems that otherwise would be difficultto solve. Here we consider special class of symmetries of Hamilton's equationcalled non-Noether symmetries. Such a symmetries appear to be closely related tomany geometric concepts used in Hamiltonian dynamics including bi-Hamiltonian structures,Frölicher-Nijenhuis operators, Lax pairs and bicomplexes.</div><div class="paragraph">Before we proceedlet us recall that each vector field <math xmlns="http://xml-maiden.com">E</math> on the phase space generatesthe one-parameter continuous group of transformations<formula xmlns="http://xml-maiden.com">g<sub>z</sub> = e<sup>zL<sub>E</sub></sup></formula> (here <math xmlns="http://xml-maiden.com">L</math> denotes Lie derivative)that acts on the observables as follows<formula xml:id="e28" xmlns="http://xml-maiden.com">g<sub>z</sub>(f) = e<sup>zL<sub>E</sub></sup>(f) =f + zL<sub>E</sub>f + ½(zL<sub>E</sub>)<sup>2</sup>f + ⋯</formula>Such a group of transformation is called symmetry of Hamilton's equation <a href="#e24">(24)</a>if it commutes with time evolution operator<formula xml:id="e29" xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction> g<sub>z</sub>(f)= g<sub>z</sub>(<fraction><num>d</num><den>dt</den></fraction>f)</formula>in terms of the vector fields this condition means that the generator<math xmlns="http://xml-maiden.com">E</math> of the group <math xmlns="http://xml-maiden.com">g<sub>z</sub></math> commutes with the vector field<math xmlns="http://xml-maiden.com">W(h) = {h , }</math>, i. e.<formula xml:id="e30" xmlns="http://xml-maiden.com">[E , W(h)] = 0.</formula> However we would like to consider more generalcase where <math xmlns="http://xml-maiden.com">E</math> is time dependent vector field on phase space. In this case<a href="#e30">(30)</a> should be replaced with<formula xml:id="e31" xmlns="http://xml-maiden.com"><fraction><num>∂</num><den>∂t</den></fraction>E = [E , W(h)].</formula></div><div class="paragraph">Further one should distinguish between groups of symmetry transformations generated by Hamiltonian,locally Hamiltonian and non-Hamiltonian vector fields. First kind of symmetriesare known as Noether symmetries and are widely used in Hamiltonian dynamics due to theirtight connection with conservation laws. Second group of symmetries is rarely used. While third group of symmetries that further will be referredas non-Noether symmetries seems to play important role in integrability issues due totheir remarkable relationship with bi-Hamiltonian structures andFrölicher-Nijenhuis operators. Thus if in addition to <a href="#e30">(30)</a> thevector field <math xmlns="http://xml-maiden.com">E</math> does not preserve Poisson bivector field <math xmlns="http://xml-maiden.com">[E , W] ≠ 0</math> then<math xmlns="http://xml-maiden.com">g<sub>z</sub></math> is called non-Noether symmetry.</div><div class="paragraph">Now let us focus on non-Noether symmetries. We would like to show that the presence ofsuch a symmetry essentially enriches the geometry of the phase spaceand under the certain conditions can ensure integrability of the dynamical system.Before we proceed let us recall that the non-Noether symmetry leads to a number ofintegrals of motion. More precisely therelationship between non-Noether symmetries and the conservation laws is described bythe following theorem. This theorem was proposed by Lutzky in <a href="#r51">[51]</a>.Here it is reformulated in terms of Poisson bivector field.</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">(M , h)</math> be regular Hamiltonian system on the <math xmlns="http://xml-maiden.com">2n</math>-dimensionalPoisson manifold <math xmlns="http://xml-maiden.com">M</math>. Then, if the vector field <math xmlns="http://xml-maiden.com">E</math> generatesnon-Noether symmetry, the functions<formula xml:id="e32" xmlns="http://xml-maiden.com">Y<sup>(k)</sup> = <fraction><num>Ŵ<sup>k</sup> ∧ W<sup>n − k</sup></num><den>W<sup>n</sup></den></fraction>           k = 1,2, ... n</formula>where <math xmlns="http://xml-maiden.com">Ŵ = [E , W]</math>, are integrals of motion. </div><div class="proof"><strong class="cap">proof. </strong>By the definition<formula xmlns="http://xml-maiden.com">Ŵ<sup>k</sup> ∧ W<sup>n − k</sup> = Y<sup>(k)</sup>W<sup>n</sup>.</formula>(definition is correct since the space of <math xmlns="http://xml-maiden.com">2n</math> degree multivector fields on <math xmlns="http://xml-maiden.com">2n</math>degree manifold is one dimensional).Let us take time derivative of this expression along the vector field <math xmlns="http://xml-maiden.com">W(h)</math>,<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Ŵ<sup>k</sup> ∧ W<sup>n − k</sup> =(<fraction><num>d</num><den>dt</den></fraction>Y<sup>(k)</sup>)W<sup>n</sup>+ Y<sup>(k)</sup>[W(h) , W<sup>n</sup>]</formula>or<formula xml:id="e35" xmlns="http://xml-maiden.com">k(<fraction><num>d</num><den>dt</den></fraction>Ŵ) ∧ Ŵ<sup>k − 1</sup> ∧ W<sup>n − k</sup>+ (n − k)[W(h) , W] ∧ Ŵ<sup>k</sup> ∧ W<sup>n − k − 1</sup> <line/>= (<fraction><num>d</num><den>dt</den></fraction>Y<sup>(k)</sup>)W<sup>n</sup>+ nY<sup>(k)</sup>[W(h) , W] ∧ W<sup>n − 1</sup></formula>but according to the Liouville theorem the Hamiltonian vector field preserves <math xmlns="http://xml-maiden.com">W</math> i. e.<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>W = [W(h) , W] = 0</formula>hence, by taking into account that<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>E= <fraction><num>∂</num><den>∂t</den></fraction>E + [W(h) , E] = 0</formula> we get<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Ŵ =<fraction><num>d</num><den>dt</den></fraction>[E , W] = [<fraction><num>d</num><den>dt</den></fraction>E , W] + [E[W(h) , W]] = 0.</formula>and as a result <a href="#e35">(35)</a> yields<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Y<sup>(k)</sup>W<sup>n</sup> = 0</formula>but since the dynamical system is regular (<math xmlns="http://xml-maiden.com">W<sup>n</sup> ≠ 0</math>)we obtain that the functions <math xmlns="http://xml-maiden.com">Y<sup>(k)</sup></math> are integrals of motion.</div><div class="remark"><strong class="cap">remark. </strong> Instead of conserved quantities<math xmlns="http://xml-maiden.com">Y<sup>(1)</sup> ... Y<sup>(n)</sup></math>, thesolutions <math xmlns="http://xml-maiden.com">c<sub>1</sub> ... c<sub>n</sub></math> of the secular equation<formula xml:id="e40" xmlns="http://xml-maiden.com">(Ŵ − cW)<sup>n</sup> = 0</formula>can be associated with the generator of symmetry.By expanding expression <a href="#e40">(40)</a> it is easy to verify that the conservation laws<math xmlns="http://xml-maiden.com">Y<sup>(k)</sup></math> can be expressed in terms of the integrals of motion<math xmlns="http://xml-maiden.com">c<sub>1</sub> ... c<sub>n</sub></math> in the following way<formula xml:id="e41" xmlns="http://xml-maiden.com">Y<sup>(k)</sup> = <fraction><num>(n − k)! k!</num><den>n!</den></fraction> <under><ope>∑</ope><sat>m<sub>s</sub> &gt; m<sub>t</sub></sat></under> c<sub>m<sub>1</sub></sub>c<sub>m<sub>2</sub></sub> ⋯ c<sub>m<sub>k</sub></sub></formula>Note also that conservation laws <math xmlns="http://xml-maiden.com">Y<sup>(k)</sup></math> can be also defined by means ofsymplectic form <math xmlns="http://xml-maiden.com">ω</math> using the following formula<formula xml:id="e42" xmlns="http://xml-maiden.com">Y<sup>(k)</sup> = <fraction><num>(L<sub>E</sub>ω)<sup>k</sup> ∧ ω<sup>n − k</sup></num><den>ω<sup>n</sup></den></fraction>       k = 1,2, ... n</formula>Conservation laws <math xmlns="http://xml-maiden.com">c<sub>1</sub> ... c<sub>n</sub></math> can be also derived fromthe secular equation<formula xml:id="e43" xmlns="http://xml-maiden.com">(L<sub>E</sub>ω − cω)<sup>n</sup> = 0</formula>However all these expressions fail in case of infinite dimensional Hamiltonian systemswhere the volume form<formula xmlns="http://xml-maiden.com">Ω = ω<sup>n</sup></formula>does not exist since <math xmlns="http://xml-maiden.com">n = ∞</math>. But fortunately in these case one can define conservation laws usingalternative formula<formula xml:id="e45" xmlns="http://xml-maiden.com">C<sup>(k)</sup> = i<sub>W<sup>k</sup></sub>(L<sub>E</sub>ω)<sup>k</sup></formula>as far as it involves only finite degree differential forms<math xmlns="http://xml-maiden.com">(L<sub>E</sub>ω)<sup>k</sup></math> and well defined multivector fields<math xmlns="http://xml-maiden.com">W<sup>k</sup></math>.Note that in finite dimensional case the sequence of conservation laws <math xmlns="http://xml-maiden.com">C<sup>(k)</sup></math>is related to families of conservation laws <math xmlns="http://xml-maiden.com">Y<sup>(k)</sup></math> and <math xmlns="http://xml-maiden.com">c<sub>k</sub></math> in thefollowing way<formula xml:id="e46" xmlns="http://xml-maiden.com">C<sup>(k)</sup> =<fraction><num>(n − k)! k!</num><den>n!</den></fraction> <under><ope>∑</ope><sat>m<sub>s</sub> &gt; m<sub>t</sub></sat></under> c<sub>m<sub>1</sub></sub>c<sub>m<sub>2</sub></sub> ⋯ c<sub>m<sub>k</sub></sub>= <fraction><num>n!</num><den>(n − k)! k!</den></fraction> Y<sup>(k)</sup></formula>Note also that by taking Lie derivative of known conservation along the generator ofsymmetry <math xmlns="http://xml-maiden.com">E</math> one can construct new conservation laws<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Y = L<sub>X<sub>h</sub></sub>Y = 0 ⇒ <fraction><num>d</num><den>dt</den></fraction>L<sub>E</sub>Y = L<sub>X<sub>h</sub></sub>L<sub>E</sub>Y =L<sub>E</sub>L<sub>X<sub>h</sub></sub>Y = 0</formula>since <math xmlns="http://xml-maiden.com">[E , X<sub>h</sub>] = 0</math>.</div><div class="remark"><strong class="cap">remark. </strong> Besides continuous non-Noether symmetries generated by non-Hamiltonianvector fields one may encounter discrete non-Noether symmetries — noncannonicaltransformations that doesn't necessarily form group but commute with evolution operator<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction> g(f) = g(<fraction><num>d</num><den>dt</den></fraction>f)</formula>Such a symmetries give rise to the same conservation laws<formula xml:id="e49" xmlns="http://xml-maiden.com">Y<sup>(k)</sup> = <fraction><num>g(W)<sup>k</sup> ∧ W<sup>n − k</sup></num><den>W<sup>n</sup></den></fraction>       k = 1,2, ... n</formula></div><div class="example"><strong class="cap">example. </strong>Let <math xmlns="http://xml-maiden.com">M</math> be <math xmlns="http://xml-maiden.com">R<sup>4</sup></math> with coordinates<math xmlns="http://xml-maiden.com">z<sub>1</sub>, z<sub>2</sub>, z<sub>3</sub>, z<sub>4</sub></math> and Poisson bivector field<formula xml:id="e50" xmlns="http://xml-maiden.com">W =<fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction></formula>and let's take the following Hamiltonian<formula xml:id="e51" xmlns="http://xml-maiden.com">h =<fraction><num>1</num><den>2</den></fraction>z<sub>1</sub><sup>2</sup> +<fraction><num>1</num><den>2</den></fraction>z<sub>2</sub><sup>2</sup> + e<sup>z<sub>3</sub> − z<sub>4</sub></sup></formula>This is so called two particle non periodic Toda model.One can check that the vector field defined as<formula xmlns="http://xml-maiden.com">E = <under><over><sat>4</sat><ope>∑</ope></over><sat>s = 1</sat></under> E<sub>s</sub><fraction><num>∂</num><den>∂z<sub>s</sub></den></fraction></formula>with components<formula xml:id="e53" xmlns="http://xml-maiden.com">E<sub>1</sub> =<fraction><num>1</num><den>2</den></fraction>z<sub>1</sub><sup>2</sup> − e<sup>z<sub>3</sub> − z<sub>4</sub></sup> −<fraction><num>t</num><den>2</den></fraction>(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>3</sub> − z<sub>4</sub></sup><line/>E<sub>2</sub> =<fraction><num>1</num><den>2</den></fraction>z<sub>2</sub><sup>2</sup> + 2e<sup>z<sub>3</sub> − z<sub>4</sub></sup> +<fraction><num>t</num><den>2</den></fraction>(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>3</sub> − z<sub>4</sub></sup><line/>E<sub>3</sub> =2z<sub>1</sub> +<fraction><num>1</num><den>2</den></fraction>z<sub>2</sub> +<fraction><num>t</num><den>2</den></fraction>(z<sub>1</sub><sup>2</sup> + e<sup>z<sub>3</sub> − z<sub>4</sub></sup>)<line/>E<sub>4</sub> = z<sub>2</sub> − <fraction><num>1</num><den>2</den></fraction>z<sub>1</sub> +<fraction><num>t</num><den>2</den></fraction>(z<sub>2</sub><sup>2</sup> + e<sup>z<sub>3</sub> − z<sub>4</sub></sup>)</formula>satisfies <a href="#e31">(31)</a> condition and as a result generates symmetry of the dynamical system.The symmetry appears to be non-Noether with Schouten bracket <math xmlns="http://xml-maiden.com">[E , W]</math> equal to<formula xml:id="e54" xmlns="http://xml-maiden.com">Ŵ = [E , W] = z<sub>1</sub><fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction>+ z<sub>2</sub><fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction>+ e<sup>z<sub>3</sub> − z<sub>4</sub></sup> <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction></formula>Calculation of volume vector fields<math xmlns="http://xml-maiden.com">Ŵ<sup>k</sup> ∧ W<sup>n − k</sup></math> gives rise to<formula xmlns="http://xml-maiden.com">W ∧ W = − 2 <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction><line/>Ŵ ∧ W = − (z<sub>1</sub> + z<sub>2</sub>)<fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction><line/>Ŵ ∧ Ŵ  =− 2(z<sub>1</sub>z<sub>2</sub> − e<sup>z<sub>3</sub> − z<sub>4</sub></sup>)<fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction></formula>and the conservation laws associated with this symmetry are just<formula xml:id="e56" xmlns="http://xml-maiden.com">Y<sup>(1)</sup> = <fraction><num>Ŵ ∧ W</num><den>W ∧ W</den></fraction> =<fraction><num>1</num><den>2</den></fraction>(z<sub>1</sub> + z<sub>2</sub>)<line/>Y<sup>(2)</sup> = <fraction><num>Ŵ ∧ Ŵ</num><den>W ∧ W</den></fraction> =z<sub>1</sub>z<sub>2</sub> − e<sup>z<sub>3</sub> − z<sub>4</sub></sup></formula>It is remarkable that the same symmetry is also present in higher dimensions.For example in case where <math xmlns="http://xml-maiden.com">M</math> is <math xmlns="http://xml-maiden.com">R<sup>6</sup></math> with coordinates<formula xmlns="http://xml-maiden.com">z<sub>1</sub>, z<sub>2</sub>, z<sub>3</sub>, z<sub>4</sub>, z<sub>5</sub>, z<sub>6</sub></formula>Poisson bivector equal to<formula xml:id="e58" xmlns="http://xml-maiden.com">W = <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> + <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction></formula>and the following Hamiltonian<formula xml:id="e59" xmlns="http://xml-maiden.com">h =<fraction><num>1</num><den>2</den></fraction>z<sub>1</sub><sup>2</sup> +<fraction><num>1</num><den>2</den></fraction>z<sub>2</sub><sup>2</sup> +<fraction><num>1</num><den>2</den></fraction>z<sub>3</sub><sup>2</sup> +e<sup>z<sub>4</sub> − z<sub>5</sub></sup> +e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula>we still can construct symmetry similar to <a href="#e53">(53)</a>.More precisely the vector field defined for arbitrary function <math xmlns="http://xml-maiden.com">F</math> as <formula xmlns="http://xml-maiden.com">E = <under><over><sat>6</sat><ope>∑</ope></over><sat>s = 1</sat></under> E<sub>s</sub><fraction><num>∂</num><den>∂z<sub>s</sub></den></fraction></formula>with components specified as follows<formula xml:id="e61" xmlns="http://xml-maiden.com">E<sub>1</sub> =<fraction><num>1</num><den>2</den></fraction>z<sub>1</sub><sup>2</sup> − 2e<sup>z<sub>4</sub> − z<sub>5</sub></sup> −<fraction><num>t</num><den>2</den></fraction>(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>4</sub> − z<sub>5</sub></sup><line/>E<sub>2</sub> =<fraction><num>1</num><den>2</den></fraction>z<sub>2</sub><sup>2</sup> + 3e<sup>z<sub>4</sub> − z<sub>5</sub></sup> −e<sup>z<sub>5</sub> − z<sub>6</sub></sup> +<fraction><num>t</num><den>2</den></fraction>(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>4</sub> − z<sub>5</sub></sup><line/>E<sub>3</sub> =<fraction><num>1</num><den>2</den></fraction>z<sub>3</sub><sup>2</sup> + 2e<sup>z<sub>5</sub> − z<sub>6</sub></sup> +<fraction><num>t</num><den>2</den></fraction>(z<sub>2</sub> + z<sub>3</sub>)e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula><formula xml:id="e62" xmlns="http://xml-maiden.com">E<sub>4</sub> =3z<sub>1</sub> + <fraction><num>1</num><den>2</den></fraction>z<sub>2</sub> + <fraction><num>1</num><den>2</den></fraction>z<sub>3</sub> +<fraction><num>t</num><den>2</den></fraction>(z<sub>1</sub><sup>2</sup> + e<sup>z<sub>4</sub> − z<sub>5</sub></sup>)<line/>E<sub>5</sub> =2z<sub>2</sub> − <fraction><num>1</num><den>2</den></fraction>z<sub>1</sub> + <fraction><num>1</num><den>2</den></fraction>z<sub>3</sub> +<fraction><num>t</num><den>2</den></fraction>(z<sub>2</sub><sup>2</sup> + e<sup>z<sub>4</sub> − z<sub>5</sub></sup> +e<sup>z<sub>5</sub> − z<sub>6</sub></sup>)<line/>E<sub>6</sub> = z<sub>3</sub> − <fraction><num>1</num><den>2</den></fraction>z<sub>1</sub> − <fraction><num>1</num><den>2</den></fraction>z<sub>2</sub> +<fraction><num>t</num><den>2</den></fraction>(z<sub>3</sub><sup>2</sup> + e<sup>z<sub>5</sub> − z<sub>6</sub></sup>)</formula>satisfies <a href="#e31">(31)</a> condition and generates non-Noether symmetry of the dynamical system(three particle non periodic Toda chain).Calculation of Schouten bracket <math xmlns="http://xml-maiden.com">[E , W]</math> gives rise to expression<formula xml:id="e63" xmlns="http://xml-maiden.com">Ŵ = [E , W] = z<sub>1</sub><fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> +z<sub>2</sub><fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> +z<sub>3</sub><fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction><line/>+ e<sup>z<sub>4</sub> − z<sub>5</sub></sup> <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +e<sup>z<sub>5</sub> − z<sub>6</sub></sup> <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> + <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction></formula>Volume multivector fields<math xmlns="http://xml-maiden.com">Ŵ<sup>k</sup> ∧ W<sup>n − k</sup></math> can be calculated in the mannersimilar to <math xmlns="http://xml-maiden.com">R<sup>4</sup></math> case and give rise to the well known conservation laws ofthree particle Toda chain.<formula xml:id="e64" xmlns="http://xml-maiden.com">Y<sup>(1)</sup> = <fraction><num>1</num><den>6</den></fraction>(z<sub>1</sub> + z<sub>2</sub> + z<sub>3</sub>) =<fraction><num>Ŵ ∧ W ∧ W</num><den>W ∧ W ∧ W</den></fraction><line/>Y<sup>(2)</sup> = <fraction><num>1</num><den>3</den></fraction>(z<sub>1</sub>z<sub>2</sub> + z<sub>1</sub>z<sub>3</sub> + z<sub>2</sub>z<sub>3</sub>− e<sup>z<sub>4</sub> − z<sub>5</sub></sup> − e<sup>z<sub>5</sub> − z<sub>6</sub></sup>)= <fraction><num>Ŵ ∧ Ŵ ∧ W</num><den>W ∧ W ∧ W</den></fraction><line/>Y<sup>(3)</sup> = z<sub>1</sub>z<sub>2</sub>z<sub>3</sub> −z<sub>3</sub>e<sup>z<sub>4</sub> − z<sub>5</sub></sup> −z<sub>1</sub>e<sup>z<sub>5</sub> − z<sub>6</sub></sup> = <fraction><num>Ŵ ∧ Ŵ ∧ Ŵ</num><den>W ∧ W ∧ W</den></fraction></formula></div><h2>Non-Liouville symmetries</h2><div class="paragraph">Besides Hamiltonian dynamical systems that admit invariant symplectic form<math xmlns="http://xml-maiden.com">ω</math>, there are dynamical systems that either are not Hamiltonian oradmit Hamiltonian realization but explicit form of symplectic structure <math xmlns="http://xml-maiden.com">ω</math>is unknown or too complex. However usually such a dynamical systems possess invariant volume form<math xmlns="http://xml-maiden.com">Ω</math> which like symplectic form can be effectively used in construction ofconservation laws. Note that volume form for given manifold is arbitrary differential formof maximal degree (equal to the dimension of manifold).In case of regular Hamiltonian systems, n-th outer power of the symplectic form <math xmlns="http://xml-maiden.com">ω</math>naturally gives rise to the invariant volume form known as Liouville form<formula xmlns="http://xml-maiden.com">Ω = ω<sup>n</sup></formula>and sometimes it is easier to work with <math xmlns="http://xml-maiden.com">Ω</math> rather then with symplectic form itself.In generic Liouville dynamical system time evolution is governed by equations of motion<formula xml:id="e66" xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>f = X(f)</formula>where <math xmlns="http://xml-maiden.com">X</math> is some smooth vector field that preserves Liouville volume form<math xmlns="http://xml-maiden.com">Ω</math><formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Ω = L<sub>X</sub>Ω = 0</formula>Symmetry of equations of motion still can be defined by condition<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction> g<sub>z</sub>(f)= g<sub>z</sub>(<fraction><num>d</num><den>dt</den></fraction>f)</formula>that in terms of vector fields implies that generator of symmetry <math xmlns="http://xml-maiden.com">E</math> shouldcommute with time evolution operator <math xmlns="http://xml-maiden.com">X</math><formula xmlns="http://xml-maiden.com">[E , X] = 0</formula>Throughout this chapter symmetry will be called non-Liouville if it is not conformal symmetryof <math xmlns="http://xml-maiden.com">Ω</math>, or in other words if<formula xmlns="http://xml-maiden.com">L<sub>E</sub>Ω ≠ cΩ</formula>for any constant <math xmlns="http://xml-maiden.com">c</math>.Such a symmetries may be considered as analog of non-Noether symmetriesdefined in Hamiltonian systems and similarly to the Hamiltonian case one can tryto construct conservation laws by means of generator of symmetry <math xmlns="http://xml-maiden.com">E</math>and invariant differential form <math xmlns="http://xml-maiden.com">Ω</math>. Namely we have the followingtheorem, which is reformulation of Hojman's theorem in terms of Liouville volume form.</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">(M, X, Ω)</math> be Liouville dynamical system on the smoothmanifold <math xmlns="http://xml-maiden.com">M</math>. Then, if the vector field <math xmlns="http://xml-maiden.com">E</math> generatesnon-Liouville symmetry, the function<formula xml:id="e71" xmlns="http://xml-maiden.com">J = <fraction><num>L<sub>E</sub>Ω</num><den>Ω</den></fraction></formula>is conservation law.</div><div class="proof"><strong class="cap">proof. </strong>By the definition<formula xmlns="http://xml-maiden.com">L<sub>E</sub>Ω = JΩ.</formula>and <math xmlns="http://xml-maiden.com">J</math> is not just constant(again definition is correct since the space of volume forms is one dimensional).By taking Lie derivative of this expression along the vector field <math xmlns="http://xml-maiden.com">X</math> thatdefines time evolution we get<formula xmlns="http://xml-maiden.com">L<sub>X</sub>L<sub>E</sub>Ω = L<sub>[X , E]</sub>Ω + L<sub>E</sub>L<sub>X</sub>Ω <line/>= L<sub>X</sub>(JΩ) = (L<sub>X</sub>J)Ω + JL<sub>X</sub>Ω</formula>but since Liouville volume form is invariant <math xmlns="http://xml-maiden.com">L<sub>X</sub>Ω = 0</math> andvector field <math xmlns="http://xml-maiden.com">E</math> is generator of symmetry satisfying <math xmlns="http://xml-maiden.com">[E , X] = 0</math>commutation relation we obtain<formula xmlns="http://xml-maiden.com">(L<sub>X</sub>J)Ω = 0</formula>or<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>J = L<sub>X</sub>J = 0</formula></div><div class="remark"><strong class="cap">remark. </strong>In fact theorem is valid for larger class of symmetries. Namely one can considersymmetries with time dependent generators. Note however that in this case condition<math xmlns="http://xml-maiden.com">[E , X] = 0</math> should be replaced by<formula xmlns="http://xml-maiden.com"><fraction><num>∂</num><den>∂t</den></fraction>E = [E , X]</formula>Note also that by calculating Lie derivative of conservation law <math xmlns="http://xml-maiden.com">J</math> alonggenerator of the symmetry <math xmlns="http://xml-maiden.com">E</math> one can recover additional conservation laws<formula xmlns="http://xml-maiden.com">J<sup>(m)</sup> = (L<sub>E</sub>)<sup>m</sup>Ω</formula></div><div class="example"><strong class="cap">example. </strong>Let us consider symmetry of three particle non periodic Toda chain. This dynamical systemwith equations of motion <formula xmlns="http://xml-maiden.com">ż<sub>4</sub> = z<sub>1</sub><line/>ż<sub>5</sub> = z<sub>2</sub><line/>ż<sub>6</sub> = z<sub>3</sub></formula><formula xmlns="http://xml-maiden.com">ż<sub>1</sub> = − e<sup>z<sub>4</sub> − z<sub>5</sub></sup><line/>ż<sub>2</sub> = e<sup>z<sub>4</sub> − z<sub>5</sub></sup> − e<sup>z<sub>5</sub> − z<sub>6</sub></sup><line/>ż<sub>3</sub> = e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula>possesses invariant volume form<formula xmlns="http://xml-maiden.com">Ω = dz<sub>1</sub> ∧ dz<sub>2</sub> ∧ dz<sub>3</sub> ∧dz<sub>4</sub> ∧ dz<sub>5</sub> ∧ dz<sub>6</sub></formula>The symmetry <a href="#e61">(61)</a> is clearly non-Liouville one as far as<formula xmlns="http://xml-maiden.com">L<sub>E</sub>Ω = (z<sub>1</sub> + z<sub>2</sub> + z<sub>3</sub>) Ω</formula>and main conservation law associated with this symmetry via Theorem 2 is total momentum<formula xmlns="http://xml-maiden.com">J = <fraction><num>L<sub>E</sub>Ω</num><den>Ω</den></fraction> = z<sub>1</sub> + z<sub>2</sub> + z<sub>3</sub></formula>Other conservation laws can be recovered by taking Lie derivative of <math xmlns="http://xml-maiden.com">J</math>along generator of symmetry <math xmlns="http://xml-maiden.com">E</math>, in particular<formula xmlns="http://xml-maiden.com">J<sup>(1)</sup> = L<sub>E</sub>J =<fraction><num>1</num><den>2</den></fraction>z<sub>1</sub><sup>2</sup> +<fraction><num>1</num><den>2</den></fraction>z<sub>2</sub><sup>2</sup> +<fraction><num>1</num><den>2</den></fraction>z<sub>3</sub><sup>2</sup> +e<sup>z<sub>4</sub> − z<sub>5</sub></sup> +e<sup>z<sub>5</sub> − z<sub>6</sub></sup><line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> =<fraction><num>1</num><den>2</den></fraction> (z<sub>1</sub><sup>3</sup> + z<sub>2</sub><sup>3</sup> + z<sub>3</sub><sup>3</sup>) <line/> + <fraction><num>3</num><den>2</den></fraction> (z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>4</sub> − z<sub>5</sub></sup> +<fraction><num>3</num><den>2</den></fraction> (z<sub>2</sub> + z<sub>3</sub>)e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula></div><h2>Lax Pairs</h2><div class="paragraph">Presence of the non-Noether symmetry not only leads to a sequence of conservation laws, but alsoendows the phase space with a number of interesting geometric structures and it appears that such asymmetry is related to many important concepts used in theory of dynamical systems.One of the such concepts is Lax pair that plays quite important role in constructionof completely integrable models.Let us recall that Lax pair of Hamiltonian system on Poisson manifold <math xmlns="http://xml-maiden.com">M</math> isa pair <math xmlns="http://xml-maiden.com">(L , P)</math> of smooth functions on <math xmlns="http://xml-maiden.com">M</math> with values in someLie algebra <math xmlns="http://xml-maiden.com">g</math> such that the time evolution of <math xmlns="http://xml-maiden.com">L</math> is given byadjoint action<formula xml:id="e84" xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>L = [L , P] = − ad<sub>P</sub>L</formula>where <math xmlns="http://xml-maiden.com">[ , ]</math> is a Lie bracket on <math xmlns="http://xml-maiden.com">g</math>. It is well known that each Laxpair leads to a number of conservation laws. When <math xmlns="http://xml-maiden.com">g</math> is some matrix Lie algebrathe conservation laws are just traces of powers of <math xmlns="http://xml-maiden.com">L</math><formula xml:id="e85" xmlns="http://xml-maiden.com">I<sup>(k)</sup> =<fraction><num>1</num><den>2</den></fraction>Tr(L<sup>k</sup>)</formula>since trace is invariant under coadjoint action<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>I<sup>(k)</sup> = <fraction><num>1</num><den>2</den></fraction> <fraction><num>d</num><den>dt</den></fraction> Tr(L<sup>k</sup>) =<fraction><num>1</num><den>2</den></fraction> Tr(<fraction><num>d</num><den>dt</den></fraction>L<sup>k</sup>) = <fraction><num>k</num><den>2</den></fraction> Tr(L<sup>k − 1</sup><fraction><num>d</num><den>dt</den></fraction>L) <line/>= <fraction><num>k</num><den>2</den></fraction> Tr(L<sup>k − 1</sup>[L , P]) = <fraction><num>1</num><den>2</den></fraction> Tr([L<sup>k</sup>, P]) = 0</formula>It is remarkable that each generator of the non-Noethersymmetry canonically leads to the Lax pair of a certain type.Such a Lax pairs have definite geometric origin, their Lax matrices are formedby coefficients of invariant tangent valued 1-form on the phase space.In the local coordinates <math xmlns="http://xml-maiden.com">z<sub>s</sub></math>, where the bivector field<math xmlns="http://xml-maiden.com">W</math>, symplectic form <math xmlns="http://xml-maiden.com">ω</math> and the generatorof the symmetry <math xmlns="http://xml-maiden.com">E</math> have the following form<formula xmlns="http://xml-maiden.com">W = <under><ope>∑</ope><sat>rs</sat></under> W<sub>rs</sub><fraction><num>∂</num><den>∂z<sub>r</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>r</sub></den></fraction>      ω = <under><ope>∑</ope><sat>rs</sat></under> ω<sub>rs</sub>dz<sub>r</sub> ∧ dz<sub>s</sub>      E = <under><ope>∑</ope><sat>s</sat></under> E<sub>s</sub><fraction><num>∂</num><den>∂z<sub>s</sub></den></fraction></formula>corresponding Lax pair can be calculated explicitly.Namely we have the following theorem (see also <a href="#r55">[55]</a>-<a href="#r56">[56]</a>):</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">(M , h)</math> be regular Hamiltonian system on the <math xmlns="http://xml-maiden.com">2n</math>-dimensionalPoisson manifold <math xmlns="http://xml-maiden.com">M</math>.Then, if the vector field <math xmlns="http://xml-maiden.com">E</math> on <math xmlns="http://xml-maiden.com">M</math> generates the non-Noether symmetry,the following <math xmlns="http://xml-maiden.com">2n×2n</math> matrix valued functions on <math xmlns="http://xml-maiden.com">M</math><formula xml:id="e88" xmlns="http://xml-maiden.com">L<sub>ab</sub> = <under><ope>∑</ope><sat>dc</sat></under> ω<sub>ad</sub><fence>E<sub>c</sub><fraction><num>∂W<sub>db</sub></num><den>∂z<sub>c</sub></den></fraction> − W<sub>bc</sub><fraction><num>∂E<sub>d</sub></num><den>∂z<sub>c</sub></den></fraction>+ W<sub>dc</sub><fraction><num>∂E<sub>b</sub></num><den>∂z<sub>c</sub></den></fraction></fence><line/>P<sub>ab</sub> =  <under><ope>∑</ope><sat>c</sat></under><fence><fraction><num>∂W<sub>bc</sub></num><den>∂z<sub>a</sub></den></fraction>·<fraction><num>∂h</num><den>∂z<sub>c</sub></den></fraction> + W<sub>bc</sub><fraction><num>∂<sup>2</sup>h</num><den>∂z<sub>a</sub>∂z<sub>c</sub></den></fraction></fence></formula>form the Lax pair <a href="#e84">(84)</a> of the dynamical system <math xmlns="http://xml-maiden.com">(M , h)</math>.</div><div class="proof"><strong class="cap">proof. </strong>Let us consider the following operator on a space of 1-forms<formula xml:id="e89" xmlns="http://xml-maiden.com">Ŕ<sub>E</sub>(u) = Φ<sub>ω</sub>([E , Φ<sub>W</sub>(u)]) − L<sub>E</sub>u</formula>(here <math xmlns="http://xml-maiden.com">Φ<sub>W</sub></math> and <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math>are maps induced by Poisson bivector field and symplectic form).It is remarkable that <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> appears to be invariant linear operator.First of all let us show that <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> is really linear,or in other words, that for arbitrary 1-forms <math xmlns="http://xml-maiden.com">u</math> and <math xmlns="http://xml-maiden.com">v</math>and function <math xmlns="http://xml-maiden.com">f</math> operator <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> has the following properties<formula xmlns="http://xml-maiden.com">Ŕ<sub>E</sub>(u + v) = Ŕ<sub>E</sub>(u) + Ŕ<sub>E</sub>(v)</formula>and<formula xmlns="http://xml-maiden.com">Ŕ<sub>E</sub>(fu) = fŔ<sub>E</sub>(u)</formula>First property is obvious consequence of linearity of Schouten bracket, Lie derivative andmaps <math xmlns="http://xml-maiden.com">Φ<sub>W</sub></math>, <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math>.Second property can be checked directly<formula xmlns="http://xml-maiden.com">Ŕ<sub>E</sub>(fu) = Φ<sub>ω</sub>([E , Φ<sub>W</sub>(fu)]) − L<sub>E</sub>(fu) <line/>= Φ<sub>ω</sub>([E , fΦ<sub>W</sub>(u)]) − (L<sub>E</sub>f)u − fL<sub>E</sub>u <line/>= Φ<sub>ω</sub>((L<sub>E</sub>f)Φ<sub>W</sub>(u)) + Φ<sub>ω</sub>(f[E , Φ<sub>W</sub>(u)]) − (L<sub>E</sub>f)u − fL<sub>E</sub>u <line/>= L<sub>E</sub>fΦ<sub>ω</sub>Φ<sub>W</sub>(u) + fΦ<sub>ω</sub>([E , Φ<sub>W</sub>(u)]) − (L<sub>E</sub>f)u − fL<sub>E</sub>u <line/>= f(Φ<sub>ω</sub>([E , Φ<sub>W</sub>(u)]) − L<sub>E</sub>u) = fŔ<sub>E</sub>(u)</formula>as far as <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub>Φ<sub>W</sub>(u) = u</math>.Now let us check that <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> is invariant operator<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Ŕ<sub>E</sub> = L<sub>X<sub>h</sub></sub>Ŕ<sub>E</sub> =L<sub>X<sub>h</sub></sub>(Φ<sub>ω</sub>L<sub>E</sub>Φ<sub>W</sub> − L<sub>E</sub>)= Φ<sub>ω</sub>L<sub>[X<sub>h</sub> , E]</sub>Φ<sub>W</sub>− L<sub>[X<sub>h</sub>, E]</sub> = 0</formula>because, being Hamiltonian vector field, <math xmlns="http://xml-maiden.com">X<sub>h</sub></math> commutes with maps<math xmlns="http://xml-maiden.com">Φ<sub>W</sub></math>, <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math>(this is consequence of Liouville theorem) and commutes with <math xmlns="http://xml-maiden.com">E</math>as far as <math xmlns="http://xml-maiden.com">E</math> generates the symmetry <math xmlns="http://xml-maiden.com">[X<sub>h</sub>, E] = 0</math>.In the terms of the local coordinates <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> has the following form<formula xmlns="http://xml-maiden.com">Ŕ<sub>E</sub> =<under><ope>∑</ope><sat>ab</sat></under>L<sub>ab</sub> dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction></formula>and the invariance condition<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Ŕ<sub>E</sub> = L<sub>W(h)</sub>Ŕ<sub>E</sub> = 0</formula>yields<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>Ŕ<sub>E</sub> = <fraction><num>d</num><den>dt</den></fraction><under><ope>∑</ope><sat>ab</sat></under>L<sub>ab</sub> dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction><line/>= <under><ope>∑</ope><sat>ab</sat></under> (<fraction><num>d</num><den>dt</den></fraction>L<sub>ab</sub>) dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction> +<under><ope>∑</ope><sat>ab</sat></under> L<sub>ab</sub> (L<sub>W(h)</sub>dz<sub>a</sub>) ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction><line/>+ <under><ope>∑</ope><sat>ab</sat></under> L<sub>ab</sub> dz<sub>a</sub> ⊗ (L<sub>W(h)</sub><fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction>) =<under><ope>∑</ope><sat>ab</sat></under> (<fraction><num>d</num><den>dt</den></fraction>L<sub>ab</sub>) dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction><line/>+ <under><ope>∑</ope><sat>abcd</sat></under>L<sub>ab</sub><fraction><num>∂W<sub>ad</sub></num><den>∂z<sub>c</sub></den></fraction>·<fraction><num>∂h</num><den>∂z<sub>d</sub></den></fraction> dz<sub>c</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction>+ <under><ope>∑</ope><sat>abcd</sat></under>L<sub>ab</sub>W<sub>ad</sub><fraction><num>∂<sup>2</sup>h</num><den>∂z<sub>c</sub>∂z<sub>d</sub></den></fraction> dz<sub>c</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction><line/>+ <under><ope>∑</ope><sat>abcd</sat></under>L<sub>ab</sub><fraction><num>∂W<sub>cd</sub></num><den>∂z<sub>b</sub></den></fraction>·<fraction><num>∂h</num><den>∂z<sub>d</sub></den></fraction>dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>c</sub></den></fraction> + <under><ope>∑</ope><sat>abcd</sat></under> L<sub>ab</sub>W<sub>cd</sub> <fraction><num>∂<sup>2</sup>h</num><den>∂z<sub>b</sub>∂z<sub>d</sub></den></fraction> dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>c</sub></den></fraction><line/>= <under><ope>∑</ope><sat>ab</sat></under><fence><fraction><num>d</num><den>dt</den></fraction>L<sub>ab</sub> + <under><ope>∑</ope><sat>c</sat></under>(P<sub>ac</sub>L<sub>cb</sub> − L<sub>ac</sub>P<sub>cb</sub>)</fence>dz<sub>a</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>b</sub></den></fraction> = 0</formula>or in matrix notations<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>L = [L , P].</formula>So, we have proved that the non-Noether symmetry canonically yields a Lax pairon the algebra of linear operators on cotangent bundle over the phase space.</div><div class="remark"><strong class="cap">remark. </strong> The conservation laws <a href="#e85">(85)</a>associated with the Lax pair <a href="#e88">(88)</a> can be expressed in terms of theintegrals of motion <math xmlns="http://xml-maiden.com">c<sub>i</sub></math> in quite simple way:<formula xml:id="e98" xmlns="http://xml-maiden.com">I<sup>(k)</sup> = <fraction><num>1</num><den>2</den></fraction> Tr(L<sup>k</sup>) = <under><ope>∑</ope><sat>s</sat></under> c<sub>s</sub><sup>k</sup></formula>This correspondence follows from the equation <a href="#e40">(40)</a>and the definition of the operator <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> <a href="#e89">(89)</a>.One can also write down recursion relation that determines conservation laws<math xmlns="http://xml-maiden.com">I<sup>(k)</sup></math> in terms of conservation laws <math xmlns="http://xml-maiden.com">C<sup>(k)</sup></math><formula xml:id="e99" xmlns="http://xml-maiden.com">I<sup>(m)</sup> + (− 1)<sup>m</sup>mC<sup>(m)</sup> +<under><over><sat>m − 1</sat><ope>∑</ope></over><sat>k = 1</sat></under>(− 1)<sup>k</sup> I<sup>(m − k)</sup>C<sup>(k)</sup> = 0</formula></div><div class="example"><strong class="cap">example. </strong>Let us calculate Lax matrix of two particle Toda chainassociated with non-Noether symmetry <a href="#e53">(53)</a>.Using <a href="#e88">(88)</a> it is easy to check that Lax matrix has eight nonzero elements<formula xml:id="e100" xmlns="http://xml-maiden.com">L =<matrix><row><cell>z<sub>1</sub></cell><cell>0</cell><cell>0</cell><cell>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup></cell></row><row><cell>0</cell><cell>z<sub>2</sub></cell><cell>e<sup>z<sub>3</sub> − z<sub>4</sub></sup></cell><cell>0</cell></row><row><cell>0</cell><cell>1</cell><cell>z<sub>1</sub></cell><cell>0</cell></row><row><cell>− 1</cell><cell>0</cell><cell>0</cell><cell>z<sub>2</sub></cell></row></matrix></formula>while matrix <math xmlns="http://xml-maiden.com">P</math> involved in Lax pair<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>L = [L , P]</formula>has the following form<formula xml:id="e102" xmlns="http://xml-maiden.com">P =<matrix><row><cell>0</cell><cell>0</cell><cell>1</cell><cell>0</cell></row><row><cell>0</cell><cell>0</cell><cell>0</cell><cell>1</cell></row><row><cell>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup></cell><cell>e<sup>z<sub>3</sub> − z<sub>4</sub></sup></cell><cell>0</cell><cell>0</cell></row><row><cell>e<sup>z<sub>3</sub> − z<sub>4</sub></sup></cell><cell>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup></cell><cell>0</cell><cell>0</cell></row></matrix></formula>The conservation laws associated with this Lax pairare total momentum and energy of two particle Toda chain<formula xml:id="e103" xmlns="http://xml-maiden.com">I<sup>(1)</sup> = <fraction><num>1</num><den>2</den></fraction> Tr(L) = z<sub>1</sub> + z<sub>2</sub><line/>I<sup>(2)</sup> = <fraction><num>1</num><den>2</den></fraction> Tr(L<sup>2</sup>) = z<sub>1</sub><sup>2</sup> + z<sub>2</sub><sup>2</sup> + 2e<sup>z<sub>3</sub> − z<sub>4</sub></sup></formula>Similarly one can construct Lax matrix of three particle Toda chain, it has 16 nonzero elements<formula xml:id="e104" xmlns="http://xml-maiden.com">L =<matrix><row><cell>z<sub>1</sub></cell><cell>0</cell><cell>0</cell><cell>0</cell><cell>− e<sup>z<sub>4</sub> − z<sub>5</sub></sup></cell><cell>0</cell></row><row><cell>0</cell><cell>z<sub>2</sub></cell><cell>0</cell><cell>e<sup>z<sub>4</sub> − z<sub>5</sub></sup></cell><cell>0</cell><cell>− e<sup>z<sub>5</sub> − z<sub>6</sub></sup></cell></row><row><cell>0</cell><cell>0</cell><cell>z<sub>3</sub></cell><cell>0</cell><cell>e<sup>z<sub>5</sub> − z<sub>6</sub></sup></cell><cell>0</cell></row><row><cell>0</cell><cell>− 1</cell><cell>− 1</cell><cell>z<sub>1</sub></cell><cell>0</cell><cell>0</cell></row><row><cell>1</cell><cell>0</cell><cell>− 1</cell><cell>0</cell><cell>z<sub>2</sub></cell><cell>0</cell></row><row><cell>1</cell><cell>1</cell><cell>0</cell><cell>0</cell><cell>0</cell><cell>z<sub>3</sub></cell></row></matrix></formula>with non-zero elements matrix <math xmlns="http://xml-maiden.com">P</math> listed below<formula xml:id="e105" xmlns="http://xml-maiden.com">P =<matrix><row><cell>0</cell><cell>0</cell><cell>0</cell><cell>1</cell><cell>0</cell><cell>0</cell></row><row><cell>0</cell><cell>0</cell><cell>0</cell><cell>0</cell><cell>1</cell><cell>0</cell></row><row><cell>0</cell><cell>0</cell><cell>0</cell><cell>0</cell><cell>0</cell><cell>1</cell></row><row><cell>− e<sup>z<sub>4</sub> − z<sub>5</sub></sup></cell><cell>e<sup>z<sub>4</sub> − z<sub>5</sub></sup></cell><cell>0</cell><cell>0</cell><cell>0</cell><cell>0</cell></row><row><cell>e<sup>z<sub>4</sub> − z<sub>5</sub></sup></cell><cell>− e<sup>z<sub>4</sub> − z<sub>5</sub></sup> − e<sup>z<sub>5</sub> − z<sub>6</sub></sup></cell><cell>e<sup>z<sub>5</sub> − z<sub>6</sub></sup></cell><cell>0</cell><cell>0</cell><cell>0</cell></row><row><cell>0</cell><cell>e<sup>z<sub>5</sub> − z<sub>6</sub></sup></cell><cell>− e<sup>z<sub>5</sub> − z<sub>6</sub></sup></cell><cell>0</cell><cell>0</cell><cell>0</cell></row></matrix></formula>Corresponding conservation laws reproduce total momentum, energy and secondHamiltonian involved in bi-Hamiltonian realization of Toda chain<formula xml:id="e106" xmlns="http://xml-maiden.com">I<sup>(1)</sup> = <fraction><num>1</num><den>2</den></fraction> Tr(L) = z<sub>1</sub> + z<sub>2</sub><line/>I<sup>(2)</sup> = <fraction><num>1</num><den>2</den></fraction> Tr(L<sup>2</sup>) = z<sub>1</sub><sup>2</sup> + z<sub>2</sub><sup>2</sup> + z<sub>3</sub><sup>2</sup> +2e<sup>z<sub>4</sub> − z<sub>5</sub></sup> + 2e<sup>z<sub>5</sub> − z<sub>6</sub></sup><line/>I<sup>(3)</sup> = <fraction><num>1</num><den>2</den></fraction> Tr(L<sup>3</sup>) =z<sub>1</sub><sup>3</sup> + z<sub>2</sub><sup>3</sup> + z<sub>3</sub><sup>3</sup> +3(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>4</sub> − z<sub>5</sub></sup> +3(z<sub>2</sub> + z<sub>3</sub>)e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula></div><h2>Involutivity of conservation laws</h2><div class="paragraph">Now let us focus on the integrability issues. We know that<math xmlns="http://xml-maiden.com">n</math> integrals of motion are associated with each generator of non-Noethersymmetry, in the same time we know that, according to the Liouville-Arnold theorem,regular Hamiltonian system <math xmlns="http://xml-maiden.com">(M, h)</math> on <math xmlns="http://xml-maiden.com">2n</math> dimensional symplectic manifold<math xmlns="http://xml-maiden.com">M</math> is completely integrable (can be solved completely) if it admits<math xmlns="http://xml-maiden.com">n</math> functionally independent integrals of motion in involution.One can understand functional independence of set of conservation laws<math xmlns="http://xml-maiden.com">c<sub>1</sub>, c<sub>2</sub> ... c<sub>n</sub></math> aslinear independence of either differentials of conservation laws<math xmlns="http://xml-maiden.com">dc<sub>1</sub>, dc<sub>2</sub> ... dc<sub>n</sub></math> orcorresponding Hamiltonian vector fields<math xmlns="http://xml-maiden.com">X<sub>c<sub>1</sub></sub>, X<sub>c<sub>2</sub></sub> ... X<sub>c<sub>n</sub></sub></math>.Strictly speaking we can say that conservation laws <math xmlns="http://xml-maiden.com">c<sub>1</sub>, c<sub>2</sub> ... c<sub>n</sub></math>are functionally independent if Lesbegue measure of the set of points of phase space <math xmlns="http://xml-maiden.com">M</math>where differentials <math xmlns="http://xml-maiden.com">dc<sub>1</sub>, dc<sub>2</sub> ... dc<sub>n</sub></math> become linearly dependentis zero. Involutivity of conservation laws means that all possible Poisson brackets ofthese conservation laws vanish pair wise<formula xmlns="http://xml-maiden.com">{c<sub>i</sub> , c<sub>j</sub>} = 0        i, j = 1... n</formula>In terms of the vector fields, existence of involutive family of <math xmlns="http://xml-maiden.com">n</math>functionally independent conservation laws<math xmlns="http://xml-maiden.com">c<sub>1</sub>, c<sub>2</sub> ... c<sub>n</sub></math>implies that corresponding Hamiltonian vector fields<math xmlns="http://xml-maiden.com">X<sub>c<sub>1</sub></sub>, X<sub>c<sub>2</sub></sub> ... X<sub>c<sub>n</sub></sub></math>span Lagrangian subspace (isotropic subspace of dimension <math xmlns="http://xml-maiden.com">n</math>)of tangent space (at each point of <math xmlns="http://xml-maiden.com">M</math>).Indeed, due to property <a href="#e23">(23)</a><formula xmlns="http://xml-maiden.com">{c<sub>i</sub> , c<sub>j</sub>} = ω(X<sub>c<sub>i</sub></sub> , X<sub>c<sub>j</sub></sub>) = 0</formula>thus space spanned by <math xmlns="http://xml-maiden.com">X<sub>c<sub>1</sub></sub>, X<sub>c<sub>2</sub></sub> ... X<sub>c<sub>n</sub></sub></math>is isotropic. Dimension of this space is <math xmlns="http://xml-maiden.com">n</math> so it is Lagrangian. Note also that distribution<math xmlns="http://xml-maiden.com">X<sub>c<sub>1</sub></sub>, X<sub>c<sub>2</sub></sub> ... X<sub>c<sub>n</sub></sub></math>is integrable since due to <a href="#e22">(22)</a><formula xmlns="http://xml-maiden.com">[X<sub>c<sub>i</sub></sub> , X<sub>c<sub>j</sub></sub>] = X<sub>{c<sub>i</sub> , c<sub>j</sub>}</sub> = 0</formula>and according to Frobenius theorem there exists submanifold of <math xmlns="http://xml-maiden.com">M</math> such thatdistribution <math xmlns="http://xml-maiden.com">X<sub>c<sub>1</sub></sub>, X<sub>c<sub>2</sub></sub> ... X<sub>c<sub>n</sub></sub></math> spans tangentspace of this submanifold. Thus for phase space geometry existence of complete involutive setof integrals of motion implies existence of invariant Lagrangian submanifold.</div><div class="paragraph">Now let us look at conservation laws <math xmlns="http://xml-maiden.com">Y<sup>(1)</sup>, Y<sup>(2)</sup> ... Y<sup>(n)</sup></math>associated with generator of non-Noether symmetry. Generally speaking these conservation laws might appear to be neither functionally independent nor involutive.However it is reasonable to ask the question – what condition should be satisfiedby the generator of the non-Noether symmetry to ensure the involutivity(<math xmlns="http://xml-maiden.com">{Y<sup>(k)</sup> , Y<sup>(m)</sup>} = 0</math>) of conserved quantities?In Lax theory situation is very similar — each Lax matrix leads to the set ofconservation laws but in general this set is not involutive, however in Lax theorythere is certain condition known as Classical Yang-Baxter Equation (CYBE)that being satisfied by Lax matrix ensures that conservation laws are in involution.Since involutivity of the conservation laws is closely related to the integrability,it is essential to have some analog of CYBE for the generatorof non-Noether symmetry. To address this issue we would like to propose the following theorem.</div><div class="theorem"><strong class="cap">theorem. </strong>If the vector field <math xmlns="http://xml-maiden.com">E</math> on <math xmlns="http://xml-maiden.com">2n</math>-dimensionalPoisson manifold <math xmlns="http://xml-maiden.com">M</math> satisfies the condition<formula xml:id="e110" xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</formula>and <math xmlns="http://xml-maiden.com">W</math> bivector field has maximal rank (<math xmlns="http://xml-maiden.com">W<sup>n</sup> ≠ 0</math>)then the functions <a href="#e32">(32)</a> are in involution<formula xmlns="http://xml-maiden.com">{Y<sup>(k)</sup> , Y<sup>(m)</sup>} = 0</formula></div><div class="proof"><strong class="cap">proof. </strong> First of all let us note thatthe identity <a href="#e15">(15)</a> satisfied by the Poissonbivector field <math xmlns="http://xml-maiden.com">W</math> is responsible for the Liouville theorem<formula xml:id="e112" xmlns="http://xml-maiden.com">[W , W] = 0       ⇔       L<sub>W(f)</sub>W = [W(f) , W] = 0</formula>that follows from the graded Jacoby identity satisfied by Schouten bracket.By taking the Lie derivative of the expression <a href="#e15">(15)</a>we obtain another useful identity<formula xmlns="http://xml-maiden.com">L<sub>E</sub>[W , W] = [E[W , W]] = [[E , W] W] + [W[E , W]]= 2[Ŵ , W] = 0.</formula>This identity gives rise to the following relation<formula xml:id="e114" xmlns="http://xml-maiden.com">[Ŵ , W] = 0      ⇔      [Ŵ(f) , W] = − [Ŵ , W(f)]</formula>and finally condition <a href="#e110">(110)</a> ensures third identity<formula xmlns="http://xml-maiden.com">[Ŵ , Ŵ] = 0</formula>yielding Liouville theorem for <math xmlns="http://xml-maiden.com">Ŵ</math><formula xml:id="e116" xmlns="http://xml-maiden.com">[Ŵ , Ŵ] = 0      ⇔      [Ŵ(f) , Ŵ] = 0</formula>Indeed<formula xmlns="http://xml-maiden.com">[Ŵ , Ŵ] = [[E , W]Ŵ] = [[Ŵ , E]W] <line/>= − [[E , Ŵ]W] = − [[E[E , W]]W] = 0</formula>Now let us consider two different solutions <math xmlns="http://xml-maiden.com">c<sub>i</sub> ≠ c<sub>j</sub></math>of the equation <a href="#e40">(40)</a>. By taking the Lie derivative of the equation<formula xmlns="http://xml-maiden.com">(Ŵ − c<sub>i</sub>W)<sup>n</sup> = 0</formula>along the vector fields <math xmlns="http://xml-maiden.com">W(c<sub>j</sub>)</math> and<math xmlns="http://xml-maiden.com">Ŵ(c<sub>j</sub>)</math> and using Liouville theorem for<math xmlns="http://xml-maiden.com">W</math> and <math xmlns="http://xml-maiden.com">Ŵ</math> bivectors we obtain the following relations<formula xml:id="e119" xmlns="http://xml-maiden.com">(Ŵ −c<sub>i</sub>W)<sup>n − 1</sup>(L<sub>W(c<sub>j</sub>)</sub>Ŵ− {c<sub>j</sub> , c<sub>i</sub>}W) =0,</formula>and<formula xml:id="e120" xmlns="http://xml-maiden.com">(Ŵ −c<sub>i</sub>W)<sup>n − 1</sup>(c<sub>i</sub>L<sub>Ŵ(c<sub>j</sub>)</sub>W+ {c<sub>j</sub> , c<sub>i</sub>}<sub>∗</sub>W) = 0,</formula>where<formula xmlns="http://xml-maiden.com">{c<sub>i</sub> , c<sub>j</sub>}<sub>∗</sub> =Ŵ(dc<sub>i</sub> ∧ dc<sub>j</sub>)</formula>is the Poisson bracket calculated by means of the bivector field <math xmlns="http://xml-maiden.com">Ŵ</math>.Now multiplying <a href="#e119">(119)</a> by <math xmlns="http://xml-maiden.com">c<sub>i</sub></math> subtracting <a href="#e120">(120)</a> and usingidentity <a href="#e114">(114)</a> gives rise to<formula xml:id="e122" xmlns="http://xml-maiden.com">({c<sub>i</sub> , c<sub>j</sub>}<sub>∗</sub> −c<sub>i</sub>{c<sub>i</sub> , c<sub>j</sub>})(Ŵ −c<sub>i</sub>W)<sup>n − 1</sup>W = 0</formula>Thus, either<formula xml:id="e123" xmlns="http://xml-maiden.com">{c<sub>i</sub> , c<sub>j</sub>}<sub>∗</sub> −c<sub>i</sub>{c<sub>i</sub> , c<sub>j</sub>} = 0</formula>or the volume field<math xmlns="http://xml-maiden.com">(Ŵ − c<sub>i</sub>W)<sup>n − 1</sup>W</math>vanishes. In the second case we can repeat<a href="#e119">(119)</a>-<a href="#e122">(122)</a> procedure forthe volume field<math xmlns="http://xml-maiden.com">(Ŵ − c<sub>i</sub>W)<sup>n − 1</sup>W</math>yielding after <math xmlns="http://xml-maiden.com">n</math>iterations <math xmlns="http://xml-maiden.com">W<sup>n</sup> = 0</math> that according to ourassumption (that the dynamical system is regular) is not true.As a result we arrived at <a href="#e123">(123)</a> and by the simpleinterchange of indices <math xmlns="http://xml-maiden.com">i ↔ j</math> we get<formula xml:id="e124" xmlns="http://xml-maiden.com">{c<sub>i</sub> , c<sub>j</sub>}<sub>∗</sub> −c<sub>j</sub>{c<sub>i</sub> , c<sub>j</sub>} = 0</formula>Finally by comparing <a href="#e123">(123)</a> and <a href="#e124">(124)</a> we obtain thatthe functions <math xmlns="http://xml-maiden.com">c<sub>i</sub></math> are in involution with respect to the bothPoisson structures (since <math xmlns="http://xml-maiden.com">c<sub>i</sub> ≠ c<sub>j</sub></math>)<formula xmlns="http://xml-maiden.com">{c<sub>i</sub> , c<sub>j</sub>}<sub>∗</sub> ={c<sub>i</sub> , c<sub>j</sub>} = 0</formula>and according to <a href="#e41">(41)</a> the same is true for the integrals of motion<math xmlns="http://xml-maiden.com">Y<sup>(k)</sup></math>.</div><div class="remark"><strong class="cap">remark. </strong> Theorem 4 is useful in multidimensional dynamical systems where involutivity ofconservation laws can not be checked directly.</div><h2>Bi-Hamiltonian systems</h2><div class="paragraph">Further we will focus on non-Noether symmetries that satisfy condition <a href="#e110">(110)</a>. Besidesyielding involutive families of conservation laws, such a symmetries appear to be relatedto many known geometric structures such as bi-Hamiltonian systems <a href="#r53">[53]</a>and Frölicher-Nijenhuis operators (torsionless tangent valued differential 1-forms).The relationship between non-Noether symmetries and bi-Hamiltonian structures wasalready implicitly outlined in the proof of Theorem 4. Now let us pay more attention tothis issue.</div><div class="paragraph">Originally bi-Hamiltonian structures were introduced by F. Magri in analisys ofintegrable infinite dimensional Hamiltonian systems such as Korteweg-de Vries (KdV) andmodified Korteweg-de Vries (mKdV) hierarchies, Nonlinear Schrödinger equationand Harry Dym equation. Since that time bi-Hamiltonian formalism is effectively usedin construction of involutive families of conservation laws in integrable models</div><div class="paragraph">Generic bi-Hamiltonian structure on <math xmlns="http://xml-maiden.com">2n</math> dimensional manifold consists outof two Poisson bivector fields <math xmlns="http://xml-maiden.com">W</math> and <math xmlns="http://xml-maiden.com">Ŵ</math> satisfying certaincompatibility condition <math xmlns="http://xml-maiden.com">[Ŵ , W] = 0</math>. If, in addition, one of these bivectorfields is nondegenerate (<math xmlns="http://xml-maiden.com">W<sup>n</sup> ≠ 0</math>) then bi-Hamiltonian systemis called regular. Further we will discuss only regular bi-Hamiltonian systems.Note that each Poisson bivector field by definition satisfies condition <a href="#e15">(15)</a>. So we actuallyimpose four restrictions on bivector fields <math xmlns="http://xml-maiden.com">W</math> and <math xmlns="http://xml-maiden.com">Ŵ</math><formula xml:id="e126" xmlns="http://xml-maiden.com">[W , W] = [Ŵ , W] = [Ŵ , Ŵ] = 0</formula>and<formula xml:id="e127" xmlns="http://xml-maiden.com">W<sup>n</sup> ≠ 0</formula>During the proof of Theorem 4 we already showed that bivector fields<math xmlns="http://xml-maiden.com">W</math> and <math xmlns="http://xml-maiden.com">Ŵ = [E , W]</math> satisfy conditions <a href="#e126">(126)</a>(see <a href="#e112">(112)</a>-<a href="#e116">(116)</a>), thus we can formulate the following statement</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">(M , h)</math> be regular Hamiltonian system on the <math xmlns="http://xml-maiden.com">2n</math>-dimensionalmanifold <math xmlns="http://xml-maiden.com">M</math> endowed with regular Poisson bivector field <math xmlns="http://xml-maiden.com">W</math>.Then, if the vector field <math xmlns="http://xml-maiden.com">E</math> on <math xmlns="http://xml-maiden.com">M</math> generates the non-Noether symmetry,and satisfies condition<formula xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0,</formula>the following bivector fields on <math xmlns="http://xml-maiden.com">M</math><formula xmlns="http://xml-maiden.com">W, Ŵ = [E , W]</formula>form invariant bi-Hamiltonian system(<math xmlns="http://xml-maiden.com">[W , W] = [Ŵ , W] = [Ŵ , Ŵ] = 0</math>).</div><div class="proof"><strong class="cap">proof. </strong> See proof of Theorem 4.</div><div class="remark"><strong class="cap">remark. </strong> Bi-Hamiltonian systems obtained by taking Lie derivative of Poisson bivector along some vector field were studied in <a href="#r70">[70]</a></div><div class="example"><strong class="cap">example. </strong> One can check that the non-Noether symmetry <a href="#e53">(53)</a> satisfiescondition <a href="#e110">(110)</a> while bivector fields<formula xmlns="http://xml-maiden.com">W =<fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction></formula>and<formula xmlns="http://xml-maiden.com">Ŵ = [E , W] = z<sub>1</sub><fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction>+ z<sub>2</sub><fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction>+ e<sup>z<sub>3</sub> − z<sub>4</sub></sup> <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction></formula>form bi-Hamiltonian system <math xmlns="http://xml-maiden.com">[W , W] = [W , Ŵ] = [Ŵ , Ŵ] = 0</math>.Similarly, one can recover bi-Hamiltonian system of three particle Toda chain associatedwith symmetry <a href="#e61">(61)</a>. It is formed by bivector fields<formula xmlns="http://xml-maiden.com">W = <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> + <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction></formula>and<formula xmlns="http://xml-maiden.com">Ŵ = [E , W] = z<sub>1</sub><fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> +z<sub>2</sub><fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> +z<sub>3</sub><fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction> <line/>+ e<sup>z<sub>4</sub> − z<sub>5</sub></sup> <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +e<sup>z<sub>5</sub> − z<sub>6</sub></sup> <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> +<fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> + <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction></formula></div><div class="paragraph">In terms of differential forms bi-Hamiltonian structure is formed by couple ofclosed differential 2-forms: symplectic form <math xmlns="http://xml-maiden.com">ω</math>(such that <math xmlns="http://xml-maiden.com">dω = 0</math> and <math xmlns="http://xml-maiden.com">ω<sup>n</sup> ≠ 0</math>)and <math xmlns="http://xml-maiden.com">ω<sup>∗</sup> = L<sub>E</sub>ω</math>(clearly <math xmlns="http://xml-maiden.com">dω<sup>∗</sup> = dL<sub>E</sub>ω= L<sub>E</sub>dω = 0</math>). It is important that by taking Lie derivative ofHamilton's equation<formula xmlns="http://xml-maiden.com">i<sub>X<sub>h</sub></sub>ω + dh = 0</formula>along the generator <math xmlns="http://xml-maiden.com">E</math> of symmetry<formula xmlns="http://xml-maiden.com">L<sub>E</sub>(i<sub>X<sub>h</sub></sub>ω + dh) =i<sub>[E , X<sub>h</sub>]</sub>ω + i<sub>X<sub>h</sub></sub>L<sub>E</sub>ω + L<sub>E</sub>dh =i<sub>X<sub>h</sub></sub>ω<sup>∗</sup> + dL<sub>E</sub>h =0</formula>one obtains another Hamilton's equation<formula xmlns="http://xml-maiden.com">i<sub>X<sub>h</sub></sub>ω<sup>∗</sup> + dh<sup>∗</sup> = 0</formula>where <math xmlns="http://xml-maiden.com">h<sup>∗</sup> = L<sub>E</sub>h</math>. This is actually second Hamiltonian realizationof equations of motion and thus under certain conditions existence of non-Noether symmetrygives rise to additional presymplectic structure <math xmlns="http://xml-maiden.com">ω<sup>∗</sup></math>and additional Hamiltonian realization of the dynamical system.In many integrable models admitting bi-Hamiltonian realization (including Toda chain,Korteweg-de Vries hierarchy, Nonlinear Schrödinger equation, Broer-Kaup system andBenney system) non-Noether symmetries that are responsible for existence of bi-Hamiltonian structureshas been found and motivated further investigation of relationship betweensymmetries and bi-Hamiltonian structures. Namely it seems to be interesting to knowwhether in general case existence of bi-Hamiltonian structure is related to non-Noether symmetry.Let us consider more general case and suppose that we have couple of differential 2-forms<math xmlns="http://xml-maiden.com">ω</math> and <math xmlns="http://xml-maiden.com">ω<sup>∗</sup></math>such that<formula xmlns="http://xml-maiden.com">dω = dω<sup>∗</sup> = 0,       ω<sup>n</sup> ≠ 0</formula><formula xmlns="http://xml-maiden.com">i<sub>X<sub>h</sub></sub>ω + dh = 0</formula>and<formula xmlns="http://xml-maiden.com">i<sub>X<sub>h</sub></sub>ω<sup>∗</sup> + dh<sup>∗</sup> = 0</formula>The question is whether there exists vector field <math xmlns="http://xml-maiden.com">E</math> (generator of non-Noether symmetry)such that <math xmlns="http://xml-maiden.com">[E , X<sub>h</sub>] = 0</math> and<math xmlns="http://xml-maiden.com">ω<sup>∗</sup> = L<sub>E</sub>ω</math>.</div><div class="paragraph">The answer depends on <math xmlns="http://xml-maiden.com">ω<sup>∗</sup></math>.Namely if <math xmlns="http://xml-maiden.com">ω<sup>∗</sup></math> is exact form(there exists 1-form <math xmlns="http://xml-maiden.com">θ<sup>∗</sup></math> such that<math xmlns="http://xml-maiden.com">ω<sup>∗</sup> = dθ<sup>∗</sup></math>)then one can argue that such a vector field exists and thus anyexact bi-Hamiltonian structure is related to hidden non-Noethersymmetry. To outline proof of this statement let us introducevector field <math xmlns="http://xml-maiden.com">E<sup>∗</sup></math> defined by<formula xmlns="http://xml-maiden.com">i<sub>E<sup>∗</sup></sub>ω = θ<sup>∗</sup></formula>(such a vector field always exist because <math xmlns="http://xml-maiden.com">ω</math>is nondegenerate 2-form).By construction<formula xmlns="http://xml-maiden.com">L<sub>E<sup>∗</sup></sub> ω = ω<sup>∗</sup></formula>Indeed<formula xmlns="http://xml-maiden.com">L<sub>E<sup>∗</sup></sub>ω = di<sub>E<sup>∗</sup></sub>ω +i<sub>E<sup>∗</sup></sub>dω = dθ<sup>∗</sup> = ω<sup>∗</sup></formula>And<formula xmlns="http://xml-maiden.com">i<sub>[E<sup>∗</sup>, X<sub>h</sub>]</sub>ω =L<sub>E<sup>∗</sup></sub>(i<sub>X<sub>h</sub></sub>ω)− i<sub>X<sub>h</sub></sub>L<sub>E<sup>∗</sup></sub>ω =− d(E<sup>∗</sup>(h)− h<sup>∗</sup>) = − dh'</formula>In other words <math xmlns="http://xml-maiden.com">[X<sub>h</sub> , E<sup>∗</sup>]</math> is Hamiltonian vector field<formula xmlns="http://xml-maiden.com">[X<sub>h</sub> , E] = X<sub>h'</sub></formula>One can also construct locally Hamiltonian vector field <math xmlns="http://xml-maiden.com">X<sub>g</sub></math>,that satisfies the same commutation relation. Namely let us definefunction (in general case this can be done only locally)<formula xmlns="http://xml-maiden.com">g(z) = <under><over><sat>t</sat><ope>∫</ope></over><sat>0</sat></under> h'dt</formula>where integration along solution of Hamilton's equation, with fixed origin and end point in<math xmlns="http://xml-maiden.com">z(t) = z</math>, is assumed.And then it is easy to verify that locally Hamiltonian vector field associated with <math xmlns="http://xml-maiden.com">g(z)</math>,by construction, satisfies the same commutation relations as<math xmlns="http://xml-maiden.com">E<sup>∗</sup></math> (namely <math xmlns="http://xml-maiden.com">[X<sub>h</sub> , X<sub>g</sub>] = X<sub>h'</sub></math>).Using <math xmlns="http://xml-maiden.com">E<sup>∗</sup></math> and <math xmlns="http://xml-maiden.com">X<sub>h'</sub></math>one can construct generator of non-Noether symmetry —non-Hamiltonian vector field <math xmlns="http://xml-maiden.com">E = E<sup>∗</sup> − X<sub>g</sub></math>commuting with <math xmlns="http://xml-maiden.com">X<sub>h</sub></math> and satisfying<formula xmlns="http://xml-maiden.com">L<sub>E</sub>ω = L<sub>E<sup>∗</sup></sub>ω− L<sub>X<sub>g</sub></sub>ω = L<sub>E<sup>∗</sup></sub>ω = ω<sup>∗</sup></formula>(thanks to Liouville's theorem <math xmlns="http://xml-maiden.com">L<sub>X<sub>g</sub></sub>ω = 0</math>). So incase of regular Hamiltonian system every exact bi-Hamiltonian structure isnaturally associated with some (non-Noether) symmetry of space of solutions.In case where bi-Hamiltonian structure is not exact(<math xmlns="http://xml-maiden.com">ω<sup>∗</sup></math> is closed but not exact) then due to<formula xmlns="http://xml-maiden.com">ω<sup>∗</sup> = L<sub>E</sub>ω =di<sub>E</sub>ω + i<sub>E</sub>dω = di<sub>E</sub>ω</formula>it is clear that such a bi-Hamiltonian system is not related to symmetry.However in all known cases bi-Hamiltonian structures seem to be exact.</div><h2>Bidifferential calculi</h2><div class="paragraph">Another important concept that is often used in theory of dynamical systems and maybe related to the non-Noether symmetry is the bidifferential calculus (bicomplex approach).Recently A. Dimakis and F. Müller-Hoissenapplied bidifferential calculi to the wide range of integrable modelsincluding KdV hierarchy, KP equation, self-dual Yang-Mills equation,Sine-Gordon equation, Toda models, non-linear Schrödingerand Liouville equations. It turns out that these models can be effectivelydescribed and analyzed using the bidifferential calculi<a href="#r17">[17]</a>, <a href="#r24">[24]</a>.Here we would like to show that each generator of non-Noether symmetrysatisfying condition <math xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</math> gives rise to certainbidifferential calculus.</div><div class="paragraph">Before we proceed let us specify what kind of bidifferential calculi we plan to consider.Under the bidifferential calculus we mean the graded algebra of differential formsover the phase space<formula xml:id="e148" xmlns="http://xml-maiden.com">Ω =<under><over><sat>∞</sat><ope>∪</ope></over><sat>k = 0</sat></under>Ω<sup>(k)</sup></formula>(<math xmlns="http://xml-maiden.com">Ω<sup>(k)</sup></math> denotes the space of <math xmlns="http://xml-maiden.com">k</math>-degree differential forms)equipped with a couple of differential operators<formula xml:id="e149" xmlns="http://xml-maiden.com">d, đ : Ω<sup>(k)</sup>  →  Ω<sup>(k + 1)</sup></formula>satisfying conditions<math xmlns="http://xml-maiden.com">d<sup>2</sup> = đ<sup>2</sup> = dđ + đd = 0</math> (see <a href="#r24">[24]</a>). In other words we have two De Rhamcomplexes <math xmlns="http://xml-maiden.com">M, Ω, d</math> and <math xmlns="http://xml-maiden.com">M, Ω, đ</math>on algebra of differential forms over the phase space. And these complexes satisfycertain compatibility condition — their differentials anticommute with each other<math xmlns="http://xml-maiden.com">dđ + đd = 0</math>.Now let us focus on non-Noether symmetries.It is interesting that if generator of the non-Noether symmetry satisfiesequation <math xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</math> then we are able to construct an invariantbidifferential calculus of a certain type.This construction is summarized in the following theorem:</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">(M , h)</math> be regular Hamiltonian system on the Poisson manifold <math xmlns="http://xml-maiden.com">M</math>.Then, if the vector field <math xmlns="http://xml-maiden.com">E</math> on <math xmlns="http://xml-maiden.com">M</math> generates the non-Noether symmetryand satisfies the equation<formula xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0,</formula>the differential operators<formula xml:id="e151" xmlns="http://xml-maiden.com">du =Φ<sub>ω</sub>([W , Φ<sub>W</sub>(u)])</formula><formula xml:id="e152" xmlns="http://xml-maiden.com">đu =Φ<sub>ω</sub>([[E , W]Φ<sub>W</sub>(u)])</formula>form invariant bidifferential calculus(<math xmlns="http://xml-maiden.com">d<sup>2</sup> = đ<sup>2</sup> = dđ + đd = 0</math>)over the graded algebra of differential forms on <math xmlns="http://xml-maiden.com">M</math>.</div><div class="proof"><strong class="cap">proof. </strong> First of all we have to show that <math xmlns="http://xml-maiden.com">d</math> and <math xmlns="http://xml-maiden.com">đ</math>are really differential operators , i.e., they are linear maps from<math xmlns="http://xml-maiden.com">Ω<sup>(k)</sup></math> into<math xmlns="http://xml-maiden.com">Ω<sup>(k + 1)</sup></math>, satisfy derivation property andare nilpotent (<math xmlns="http://xml-maiden.com">d<sup>2</sup> = đ<sup>2</sup> = 0</math>).Linearity is obvious and follows from the linearity of the Schouten bracket <math xmlns="http://xml-maiden.com">[ , ]</math>and <math xmlns="http://xml-maiden.com">Φ<sub>W</sub>, Φ<sub>ω</sub></math>maps. Then, if <math xmlns="http://xml-maiden.com">u</math> is a <math xmlns="http://xml-maiden.com">k</math>-degree form<math xmlns="http://xml-maiden.com">Φ<sub>W</sub></math> maps it on <math xmlns="http://xml-maiden.com">k</math>-degree multivector field andthe Schouten brackets <math xmlns="http://xml-maiden.com">[W , Φ<sub>W</sub>(u)]</math> and<math xmlns="http://xml-maiden.com">[[E , W]Φ<sub>W</sub>(u)]</math> result the<math xmlns="http://xml-maiden.com">k + 1</math>-degree multivector fields that are mapped on <math xmlns="http://xml-maiden.com">k + 1</math>-degreedifferential forms by <math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math>.So, <math xmlns="http://xml-maiden.com">d</math> and <math xmlns="http://xml-maiden.com">đ</math>are linear maps from <math xmlns="http://xml-maiden.com">Ω<sup>(k)</sup></math> into<math xmlns="http://xml-maiden.com">Ω<sup>(k + 1)</sup></math>.Derivation property follows from the same feature of the Schouten bracket<math xmlns="http://xml-maiden.com">[ , ]</math> and linearity of<math xmlns="http://xml-maiden.com">Φ<sub>W</sub></math> and<math xmlns="http://xml-maiden.com">Φ<sub>ω</sub></math> maps.Now we have to prove the nilpotency of <math xmlns="http://xml-maiden.com">d</math> and <math xmlns="http://xml-maiden.com">đ</math>.Let us consider <math xmlns="http://xml-maiden.com">d<sup>2</sup>u</math><formula xmlns="http://xml-maiden.com">d<sup>2</sup>u =Φ<sub>ω</sub>([W , Φ<sub>W</sub>(Φ<sub>ω</sub>([W , Φ<sub>W</sub>(u)]))])= Φ<sub>ω</sub>([W[W , Φ<sub>W</sub>(u)]]) = 0</formula>as a result of the property <a href="#e112">(112)</a> and the Jacoby identity for <math xmlns="http://xml-maiden.com">[ , ]</math> bracket.In the same manner<formula xmlns="http://xml-maiden.com">đ<sup>2</sup>u =Φ<sub>ω</sub>([[W , E][[W , E]Φ<sub>W</sub>(u)]]) = 0</formula>according to the property <a href="#e116">(116)</a> of<math xmlns="http://xml-maiden.com">[W , E] = Ŵ</math> and the Jacoby identity.Thus, we have proved that <math xmlns="http://xml-maiden.com">d</math> and <math xmlns="http://xml-maiden.com">đ</math> are differential operators(in fact <math xmlns="http://xml-maiden.com">d</math> is ordinary exterior differential and the expression<a href="#e151">(151)</a> is its well known representation in terms of Poisson bivector field).It remains to show that the compatibility condition <math xmlns="http://xml-maiden.com">dđ + đd = 0</math>is fulfilled. Using definitions of <math xmlns="http://xml-maiden.com">d, đ</math> and the Jacoby identity we get<formula xmlns="http://xml-maiden.com">(dđ + đd)(u) =Φ<sub>ω</sub>([[[W , E]W]Φ<sub>W</sub>(u)]) = 0</formula>as far as <a href="#e114">(114)</a> is satisfied.So, <math xmlns="http://xml-maiden.com">d</math> and <math xmlns="http://xml-maiden.com">đ</math> form the bidifferential calculus over the gradedalgebra of differential forms.It is also clear that the bidifferential calculus <math xmlns="http://xml-maiden.com">d, đ</math>is invariant, since both <math xmlns="http://xml-maiden.com">d</math> and <math xmlns="http://xml-maiden.com">đ</math> commute with time evolutionoperator <math xmlns="http://xml-maiden.com">W(h) = {h, }</math>.</div><div class="remark"><strong class="cap">remark. </strong>Conservation laws that are associated with the bidifferential calculus<a href="#e151">(151)</a> <a href="#e152">(152)</a> and form Lenard scheme (see <a href="#r24">[24]</a>): <formula xmlns="http://xml-maiden.com">(k + 1)đI<sup>(k)</sup> = kdI<sup>(k + 1)</sup></formula>coincide with the sequence of integrals of motion <a href="#e98">(98)</a>.Proof of this correspondence lays outside the scope of present manuscript,but can be done in the manner similar to <a href="#r17">[17]</a>.</div><div class="example"><strong class="cap">example. </strong>The symmetry <a href="#e53">(53)</a> endows <math xmlns="http://xml-maiden.com">R<sup>4</sup></math> with bicomplex structure<math xmlns="http://xml-maiden.com">d, đ</math> where <math xmlns="http://xml-maiden.com">d</math> is ordinary exterior derivative while <math xmlns="http://xml-maiden.com">đ</math>is defined by<formula xml:id="e157" xmlns="http://xml-maiden.com">đz<sub>1</sub> = z<sub>1</sub>dz<sub>1</sub> − e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>4</sub><line/>đz<sub>2</sub> = z<sub>2</sub>dz<sub>2</sub> + e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>3</sub><line/>đz<sub>3</sub> = z<sub>1</sub>dz<sub>3</sub> + dz<sub>2</sub><line/>đz<sub>4</sub> = z<sub>2</sub>dz<sub>4</sub> − dz<sub>1</sub></formula>and is extended to whole De Rham complex by linearity, derivation property andcompatibility property <math xmlns="http://xml-maiden.com">dđ + đd = 0</math>.By direct calculations one can verify that calculus constructed in this wayis consistent and satisfies <math xmlns="http://xml-maiden.com">đ<sup>2</sup> = 0</math> property.To illustrate technique let us explicitly check that <math xmlns="http://xml-maiden.com">đ<sup>2</sup>z<sub>1</sub> = 0</math>.Indeed<formula xmlns="http://xml-maiden.com">đ<sup>2</sup>z<sub>1</sub> = đđz<sub>1</sub> =đ(z<sub>1</sub>dz<sub>1</sub> − e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>4</sub>) <line/>= đz<sub>1</sub> ∧ dz<sub>1</sub> + z<sub>1</sub>đdz<sub>1</sub>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đz<sub>3</sub> ∧ dz<sub>4</sub>+ e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đz<sub>4</sub> ∧ dz<sub>4</sub>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đdz<sub>4</sub> <line/>= đz<sub>1</sub> ∧ dz<sub>1</sub> − z<sub>1</sub>dđz<sub>1</sub>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đz<sub>3</sub> ∧ dz<sub>4</sub>+ e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đz<sub>4</sub> ∧ dz<sub>4</sub>+ e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dđz<sub>4</sub> = 0</formula>Because of properties<formula xmlns="http://xml-maiden.com">đz<sub>1</sub> ∧ dz<sub>1</sub> =e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>1</sub> ∧ dz<sub>4</sub>,</formula><formula xmlns="http://xml-maiden.com">− z<sub>1</sub>dđz<sub>1</sub> =z<sub>1</sub>e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>3</sub> ∧ dz<sub>4</sub>,</formula><formula xmlns="http://xml-maiden.com">− e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đz<sub>3</sub> ∧ dz<sub>4</sub> <line/>= − z<sub>1</sub>e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>1</sub> ∧ dz<sub>4</sub>− e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>2</sub> ∧ dz<sub>4</sub>,</formula><formula xmlns="http://xml-maiden.com">e<sup>z<sub>3</sub> − z<sub>4</sub></sup>đz<sub>4</sub> ∧ dz<sub>4</sub> =e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>2</sub> ∧ dz<sub>4</sub></formula>and<formula xmlns="http://xml-maiden.com">e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dđz<sub>4</sub> =− e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>1</sub> ∧ dz<sub>4</sub></formula>Similarly one can show that<formula xmlns="http://xml-maiden.com">đ<sup>2</sup>z<sub>2</sub> = đ<sup>2</sup>z<sub>3</sub> = đ<sup>2</sup>z<sub>4</sub> = 0</formula>and thus <math xmlns="http://xml-maiden.com">đ</math> is nilpotent operator <math xmlns="http://xml-maiden.com">đ<sup>2</sup> = 0</math>.Note also that conservation laws<formula xmlns="http://xml-maiden.com">I<sup>(1)</sup> = z<sub>1</sub> + z<sub>2</sub><line/>I<sup>(2)</sup> = z<sub>1</sub><sup>2</sup> + z<sub>2</sub><sup>2</sup> + 2e<sup>z<sub>3</sub> − z<sub>4</sub></sup></formula>form the simplest Lenard scheme<formula xmlns="http://xml-maiden.com">2đI<sup>(1)</sup> = dI<sup>(2)</sup></formula>Similarly one can construct bidifferential calculus associated with non-Noethersymmetry <a href="#e61">(61)</a> of three particle Toda chain. In this case <math xmlns="http://xml-maiden.com">đ</math>can be defined by<formula xml:id="e167" xmlns="http://xml-maiden.com">đz<sub>1</sub> = z<sub>1</sub>dz<sub>1</sub> − e<sup>z<sub>4</sub> − z<sub>5</sub></sup>dz<sub>5</sub><line/>đz<sub>2</sub> = z<sub>2</sub>dz<sub>2</sub> + e<sup>z<sub>4</sub> − z<sub>5</sub></sup>dz<sub>4</sub>− e<sup>z<sub>5</sub> − z<sub>6</sub></sup>dz<sub>6</sub><line/>đz<sub>3</sub> = z<sub>3</sub>dz<sub>3</sub> + e<sup>z<sub>5</sub> − z<sub>6</sub></sup>dz<sub>5</sub><line/>đz<sub>4</sub> = z<sub>1</sub>dz<sub>4</sub> − dz<sub>2</sub> − dz<sub>3</sub><line/>đz<sub>5</sub> = z<sub>2</sub>dz<sub>5</sub> + dz<sub>1</sub> − dz<sub>3</sub><line/>đz<sub>6</sub> = z<sub>3</sub>dz<sub>6</sub> + dz<sub>1</sub> + dz<sub>2</sub></formula>and as in case of two particle Toda itcan be extended to whole De Rham complex by linearity, derivation property andcompatibility property <math xmlns="http://xml-maiden.com">dđ + đd = 0</math>.One can check that conservation laws of Toda chain<formula xmlns="http://xml-maiden.com">I<sup>(1)</sup> = z<sub>1</sub> + z<sub>2</sub><line/>I<sup>(2)</sup> =z<sub>1</sub><sup>2</sup> + z<sub>2</sub><sup>2</sup> + z<sub>3</sub><sup>2</sup> +2e<sup>z<sub>4</sub> − z<sub>5</sub></sup> + 2e<sup>z<sub>5</sub> − z<sub>6</sub></sup><line/>I<sup>(3)</sup> =z<sub>1</sub><sup>3</sup> + z<sub>2</sub><sup>3</sup> + z<sub>3</sub><sup>3</sup> +3(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>4</sub> − z<sub>5</sub></sup> +3(z<sub>2</sub> + z<sub>3</sub>)e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula>form Lenard scheme<formula xmlns="http://xml-maiden.com">2đI<sup>(1)</sup> = dI<sup>(2)</sup></formula><formula xmlns="http://xml-maiden.com">3đI<sup>(2)</sup> = 2dI<sup>(3)</sup></formula></div><h2>Frölicher-Nijenhuis geometry</h2><div class="paragraph">Finally we would like to reveal some features of the operator<math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math><a href="#e89">(89)</a> and to show how Frölicher-Nijenhuis geometry arises inHamiltonian system that possesses certain non-Noether symmetry.From the geometric properties of the tangent valued forms we knowthat the traces of powers of a linear operator <math xmlns="http://xml-maiden.com">F</math>on tangent bundle are in involution whenever its Frölicher-Nijenhuis torsion<math xmlns="http://xml-maiden.com">T(F)</math> vanishes, i. e. whenever for arbitrary vector fields <math xmlns="http://xml-maiden.com">X,Y</math> the condition<formula xmlns="http://xml-maiden.com">T(F)(X , Y) = [FX , FY] −F([FX , Y] + [X , FY] − F[X , Y]) = 0</formula>is satisfied.Torsionless forms are also called Frölicher-Nijenhuis operators and are widely used intheory of integrable models, where they play role of recursion operators and are usedin construction of involutive family of conservation laws.We would like to show that each generator of non-Noether symmetry satisfying equation<math xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</math>canonically leads to invariant Frölicher-Nijenhuis operator on tangentbundle over the phase space. This operator can be expressed in terms of generator of symmetryand isomorphism defined by Poisson bivector field. Strictly speaking we have the following theorem.</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">(M , h)</math> be regular Hamiltonian system on the Poisson manifold <math xmlns="http://xml-maiden.com">M</math>.If the vector field <math xmlns="http://xml-maiden.com">E</math> on <math xmlns="http://xml-maiden.com">M</math> generates the non-Noether symmetryand satisfies the equation<formula xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</formula>then the linear operator, defined forevery vector field <math xmlns="http://xml-maiden.com">X</math> by equation<formula xml:id="e173" xmlns="http://xml-maiden.com">R<sub>E</sub>(X) =Φ<sub>W</sub>(L<sub>E</sub>Φ<sub>ω</sub>(X))− [E , X]</formula>is invariant Frölicher-Nijenhuis operator on <math xmlns="http://xml-maiden.com">M</math>.</div><div class="proof"><strong class="cap">proof. </strong>Invariance of <math xmlns="http://xml-maiden.com">R<sub>E</sub></math> follows from the invariance of the<math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math> defined by <a href="#e89">(89)</a>(note that for arbitrary 1-form vector field <math xmlns="http://xml-maiden.com">u</math> and vector field <math xmlns="http://xml-maiden.com">X</math>contraction <math xmlns="http://xml-maiden.com">i<sub>X</sub>u</math> has the property<math xmlns="http://xml-maiden.com">i<sub>R<sub>E</sub>X</sub>u =i<sub>X</sub>Ŕ<sub>E</sub>u</math>,so <math xmlns="http://xml-maiden.com">R<sub>E</sub></math> is actually transposed to<math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math>).It remains to show that the condition <a href="#e110">(110)</a> ensures vanishing of theFrölicher-Nijenhuis torsion <math xmlns="http://xml-maiden.com">T(R<sub>E</sub>)</math> of<math xmlns="http://xml-maiden.com">R<sub>E</sub></math>, i.e. for arbitrary vector fields <math xmlns="http://xml-maiden.com">X, Y</math> we must get<formula xml:id="e174" xmlns="http://xml-maiden.com">T(R<sub>E</sub>)(X , Y) = [R<sub>E</sub>(X) , R<sub>E</sub>(Y)] −R<sub>E</sub>([R<sub>E</sub>(X) , Y]<line/> + [X , R<sub>E</sub>(Y)] − R<sub>E</sub>([X , Y])) = 0</formula>First let us introduce the following auxiliary 2-forms<formula xml:id="e175" xmlns="http://xml-maiden.com">ω = Φ<sub>ω</sub>(W),       ω<sup>∗</sup> = Ŕ<sub>E</sub>ω       ω<sup>∗∗</sup> = Ŕ<sub>E</sub>ω<sup>∗</sup></formula>Using the realization <a href="#e151">(151)</a> of the differential <math xmlns="http://xml-maiden.com">d</math>and the property <a href="#e15">(15)</a> yields<formula xmlns="http://xml-maiden.com">dω = Φ<sub>ω</sub>([W , W]) = 0</formula>Similarly, using the property <a href="#e114">(114)</a> we obtain<formula xmlns="http://xml-maiden.com">dω<sup>∗</sup> =dΦ<sub>ω</sub>([E , W]) − dL<sub>E</sub>ω =Φ<sub>ω</sub>([[E , W]W]) −L<sub>E</sub>dω = 0</formula>And finally, taking into account that<math xmlns="http://xml-maiden.com">ω<sup>∗</sup> = 2Φ<sub>ω</sub>([E , W])</math>and using the condition <a href="#e110">(110)</a>, we get<formula xmlns="http://xml-maiden.com">dω<sup>∗∗</sup> =2Φ<sub>ω</sub>([[E[E , W]]W])− 2dL<sub>E</sub>ω<sup>∗</sup> =− 2L<sub>E</sub>dω<sup>∗</sup> = 0</formula>So the differential forms<math xmlns="http://xml-maiden.com">ω, ω<sup>∗</sup>, ω<sup>∗∗</sup></math>are closed<formula xml:id="e179" xmlns="http://xml-maiden.com">dω = dω<sup>∗</sup> = dω<sup>∗∗</sup> = 0</formula>Now let us consider the contraction of <math xmlns="http://xml-maiden.com">T(R<sub>E</sub>)</math> and <math xmlns="http://xml-maiden.com">ω</math>.<formula xml:id="e180" xmlns="http://xml-maiden.com">i<sub>T(R<sub>E</sub>)(X , Y)</sub>ω =i<sub>[R<sub>E</sub>X , R<sub>E</sub>Y]</sub>ω −i<sub>[R<sub>E</sub>X , Y]</sub>ω<sup>∗</sup> −i<sub>[X , R<sub>E</sub>Y]</sub>ω<sup>∗</sup> +i<sub>[X , Y]</sub>ω<sup>∗∗</sup><line/>=L<sub>R<sub>E</sub>X</sub>i<sub>Y</sub>ω<sup>∗</sup> −i<sub>R<sub>E</sub>Y</sub>L<sub>X</sub>ω<sup>∗</sup> −L<sub>R<sub>E</sub>X</sub>i<sub>Y</sub>ω<sup>∗</sup> +i<sub>Y</sub>L<sub>R<sub>E</sub>X</sub>ω<sup>∗</sup> −L<sub>X</sub>i<sub>R<sub>E</sub>Y</sub>ω<sup>∗</sup> +i<sub>R<sub>E</sub>Y</sub>L<sub>X</sub>ω<sup>∗</sup> +i<sub>[X , Y]</sub>ω<sup>∗∗</sup><line/>= i<sub>Y</sub>L<sub>X</sub>ω<sup>∗∗</sup> −L<sub>X</sub>i<sub>Y</sub>ω<sup>∗∗</sup> +i<sub>[X , Y]</sub>ω<sup>∗∗</sup> = 0</formula>where we used <a href="#e175">(175)</a> <a href="#e179">(179)</a>,the property <formula xmlns="http://xml-maiden.com">L<sub>X</sub>i<sub>Y</sub>ω =i<sub>Y</sub>L<sub>X</sub>ω + i<sub>[X , Y]</sub>ω</formula>of the Lie derivative and the relations of the following type<formula xmlns="http://xml-maiden.com">L<sub>R<sub>E</sub>X</sub>ω =di<sub>R<sub>E</sub>X</sub>ω + i<sub>R<sub>E</sub>X</sub>dω= di<sub>X</sub>ω<sup>∗</sup><line/>= L<sub>X</sub>ω<sup>∗</sup> −i<sub>X</sub>dω<sup>∗</sup> = L<sub>X</sub>ω<sup>∗</sup></formula>So we proved that for arbitrary vector fields <math xmlns="http://xml-maiden.com">X, Y</math>the contraction of <math xmlns="http://xml-maiden.com">T(R<sub>E</sub>)(X , Y)</math> and <math xmlns="http://xml-maiden.com">ω</math> vanishes.But since <math xmlns="http://xml-maiden.com">W</math> bivector is non-degenerate(<math xmlns="http://xml-maiden.com">W<sup>n</sup> ≠ 0</math>), its counter image<formula xmlns="http://xml-maiden.com">ω = Φ<sub>ω</sub>(W)</formula>is also non-degenerate and vanishing of the contraction <a href="#e180">(180)</a>implies that the torsion <math xmlns="http://xml-maiden.com">T(R<sub>E</sub>)</math> itself is zero.So we get<formula xmlns="http://xml-maiden.com">T(R<sub>E</sub>)(X , Y) = [R<sub>E</sub>(X) , R<sub>E</sub>(Y)] −R<sub>E</sub>([R<sub>E</sub>(X) , Y] <line/>+ [X , R<sub>E</sub>(Y)] − R<sub>E</sub>([X , Y])) = 0</formula></div><div class="example"><strong class="cap">example. </strong> The operator <math xmlns="http://xml-maiden.com">R<sub>E</sub></math> associated with non-Noethersymmetry <a href="#e53">(53)</a> reproduces well known Frölicher-Nijenhuis operator<formula xmlns="http://xml-maiden.com">R<sub>E</sub> =z<sub>1</sub>dz<sub>1</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction> −dz<sub>1</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> +z<sub>2</sub>dz<sub>2</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +dz<sub>2</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> <line/>+ z<sub>1</sub>dz<sub>3</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> +e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>3</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +z<sub>2</sub>dz<sub>4</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction> −e<sup>z<sub>3</sub> − z<sub>4</sub></sup>dz<sub>4</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction></formula>(compare with <a href="#r30">[30]</a>).The operator <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math>plays the role of recursion operator for conservation laws<formula xmlns="http://xml-maiden.com">I<sup>(1)</sup> = z<sub>1</sub> + z<sub>2</sub><line/>I<sup>(2)</sup> = z<sub>1</sub><sup>2</sup> + z<sub>2</sub><sup>2</sup> + 2e<sup>z<sub>3</sub> − z<sub>4</sub></sup></formula>Indeed one can check that<formula xmlns="http://xml-maiden.com">2Ŕ<sub>E</sub>(dI<sup>(1)</sup>) = dI<sup>(2)</sup></formula>Similarly using non-Noether symmetry <a href="#e61">(61)</a> one can construct recursion operator ofthree particle Toda chain<formula xmlns="http://xml-maiden.com">R<sub>E</sub> = z<sub>1</sub>dz<sub>1</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction>− e<sup>z<sub>4</sub> − z<sub>5</sub></sup>dz<sub>5</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>1</sub></den></fraction><line/>+  z<sub>2</sub>dz<sub>2</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction> +e<sup>z<sub>4</sub> − z<sub>5</sub></sup>dz<sub>4</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction><line/>− e<sup>z<sub>5</sub> − z<sub>6</sub></sup>dz<sub>6</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>2</sub></den></fraction>+z<sub>3</sub>dz<sub>3</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction> +e<sup>z<sub>5</sub> − z<sub>6</sub></sup>dz<sub>5</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>3</sub></den></fraction><line/>+ z<sub>1</sub>dz<sub>4</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction>− dz<sub>2</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction>− dz<sub>3</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>4</sub></den></fraction><line/>+ z<sub>2</sub>dz<sub>5</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction> + dz<sub>1</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction>− dz<sub>3</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>5</sub></den></fraction><line/>+ z<sub>3</sub>dz<sub>6</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction> +dz<sub>1</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction> + dz<sub>2</sub> ⊗ <fraction><num>∂</num><den>∂z<sub>6</sub></den></fraction></formula>and as in case of two particle Toda chain, operator <math xmlns="http://xml-maiden.com">Ŕ<sub>E</sub></math>appears to be recursion operator for conservation laws<formula xmlns="http://xml-maiden.com">I<sup>(1)</sup> = z<sub>1</sub> + z<sub>2</sub><line/>I<sup>(2)</sup> =z<sub>1</sub><sup>2</sup> + z<sub>2</sub><sup>2</sup> + z<sub>3</sub><sup>2</sup> +2e<sup>z<sub>4</sub> − z<sub>5</sub></sup> + 2e<sup>z<sub>5</sub> − z<sub>6</sub></sup><line/>I<sup>(3)</sup> =z<sub>1</sub><sup>3</sup> + z<sub>2</sub><sup>3</sup> + z<sub>3</sub><sup>3</sup><line/>+ 3(z<sub>1</sub> + z<sub>2</sub>)e<sup>z<sub>4</sub> − z<sub>5</sub></sup> + 3(z<sub>2</sub> + z<sub>3</sub>)e<sup>z<sub>5</sub> − z<sub>6</sub></sup></formula>and fulfills the following recursion condition<formula xmlns="http://xml-maiden.com">dI<sup>(3)</sup> = 3Ŕ<sub>E</sub>(dI<sup>(2)</sup>) =6(Ŕ<sub>E</sub>)<sup>2</sup>(dI<sup>(1)</sup>)</formula></div><h2>One-parameter families of conservation laws</h2><div class="paragraph">One-parameter group of transformations <math xmlns="http://xml-maiden.com">g<sub>z</sub></math>defined by <a href="#e28">(28)</a> naturally acts on algebra of integrals of motion.Namely for each conservation law<formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>J = 0</formula>one can define one-parameter family of conserved quantities <math xmlns="http://xml-maiden.com">J(z)</math>by applying group of transformations <math xmlns="http://xml-maiden.com">g<sub>z</sub></math> to <math xmlns="http://xml-maiden.com">J</math><formula xmlns="http://xml-maiden.com">J(z) = g<sub>z</sub>(J) = e<sup>zL<sub>E</sub></sup>J =J + zL<sub>E</sub>J + ½(zL<sub>E</sub>)<sup>2</sup>J + ...</formula>Property <a href="#e29">(29)</a> ensures that <math xmlns="http://xml-maiden.com">J(z)</math> is conserved for arbitrary valuesof parameter <math xmlns="http://xml-maiden.com">z</math><formula xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>J(z) =<fraction><num>d</num><den>dt</den></fraction>g<sub>z</sub>(J) =g<sub>z</sub>(<fraction><num>d</num><den>dt</den></fraction>J) = 0</formula>and thus each conservation law gives rise to whole family of conservedquantities that form orbit of group of transformations <math xmlns="http://xml-maiden.com">g<sub>a</sub></math>.</div><div class="paragraph">Such an orbit <math xmlns="http://xml-maiden.com">J(z)</math> is called involutive if conservation laws that formit are in involution<formula xmlns="http://xml-maiden.com">{J(z<sub>1</sub>) , J(z<sub>2</sub>)} = 0</formula>(for arbitrary values of parameters <math xmlns="http://xml-maiden.com">z<sub>1</sub>, z<sub>2</sub></math>). On <math xmlns="http://xml-maiden.com">2n</math> dimensionalsymplectic manifold each involutive family that contains <math xmlns="http://xml-maiden.com">n</math> functionally independentintegrals of motion naturally gives rise to integrable system (due to Liouville-Arnold theorem).So in order to identify those orbits that may be related to integrable models itis important to know how involutivity of family of conserved quantities <math xmlns="http://xml-maiden.com">J(z)</math>is related to properties of initial conserved quantity <math xmlns="http://xml-maiden.com">J(0) = J</math> and nature ofgenerator <math xmlns="http://xml-maiden.com">E</math> of group <math xmlns="http://xml-maiden.com">g<sub>z</sub> = e<sup>zL<sub>E</sub></sup></math>.In other words we would like to know what condition must be satisfied by generator ofsymmetry <math xmlns="http://xml-maiden.com">E</math> and integral of motion <math xmlns="http://xml-maiden.com">J</math> to ensure that<math xmlns="http://xml-maiden.com">{J(z<sub>1</sub>) , J(z<sub>2</sub>)} = 0</math>. To address this issue and to describe class of vector fieldsthat possess nontrivial involutive orbits we would like to propose the followingtheorem</div><div class="theorem"><strong class="cap">theorem. </strong>Let <math xmlns="http://xml-maiden.com">M</math> be Poisson manifold endowed with 1-form <math xmlns="http://xml-maiden.com">s</math>such that<formula xml:id="e195" xmlns="http://xml-maiden.com">[W[W(s),W](s)] = c<sub>0</sub>[W(s)[W(s) ,W]]       (c<sub>0</sub> ≠ − 1)</formula>Then each function <math xmlns="http://xml-maiden.com">J</math> satisfying property<formula xml:id="e196" xmlns="http://xml-maiden.com">W(L<sub>W(s)</sub>dJ) = c<sub>1</sub>[W(s),W](dJ)      (c<sub>1</sub> ≠ 0)</formula>(<math xmlns="http://xml-maiden.com">c<sub>0,1</sub></math> are some constants) gives rise to involutiveset of functions<formula xml:id="e197" xmlns="http://xml-maiden.com">J<sup>(m)</sup> = (L<sub>W(s)</sub>)<sup>m</sup>J       {J<sup>(m)</sup>, J<sup>(k)</sup>} = 0</formula></div><div class="proof"><strong class="cap">proof. </strong>First let us inroduce linear operator <math xmlns="http://xml-maiden.com">R</math> on bundle of multivector fields and define itfor arbitrary multivector field <math xmlns="http://xml-maiden.com">V</math> by condition<formula xml:id="e198" xmlns="http://xml-maiden.com">R(V) = ½ ([W(s),V] − Φ<sub>W</sub>(L<sub>W(s)</sub>Φ<sub>ω</sub>(V)))</formula>Proof of linearity of this operator is identical to proof given for<a href="#e89">(89)</a> so we will skip it. Further it is clear that<formula xml:id="e199" xmlns="http://xml-maiden.com">R(W) = [W(s),W]</formula>and<formula xml:id="e200" xmlns="http://xml-maiden.com">R<sup>2</sup>(W) = R([W(s),W]) = ½([W(s)[W(s),W]] − Φ<sub>W</sub>((L<sub>W(s)</sub>)<sup>2</sup>ω))<line/>= ½(1 + c<sub>0</sub>)[W(s)[W(s),W]]</formula>where we used property<formula xmlns="http://xml-maiden.com">Φ<sub>W</sub>((L<sub>W(s)</sub>)<sup>2</sup>ω) =Φ<sub>W</sub>(L<sub>W(s)</sub>L<sub>W(s)</sub>ω) <line/>= Φ<sub>W</sub>(i<sub>W(s)</sub>dL<sub>W(s)</sub>ω) +Φ<sub>W</sub>(di<sub>W(s)</sub>L<sub>W(s)</sub>ω) <line/>= [W,Φ<sub>W</sub>(i<sub>W(s)</sub>L<sub>W(s)</sub>ω)] =[W[W(s),W](s)] = c<sub>0</sub>[W(s)[W(s),W]]</formula>At the same time by taking Lie derivative of<a href="#e199">(199)</a> along the vector field <math xmlns="http://xml-maiden.com">W(s)</math>one gets<formula xml:id="e202" xmlns="http://xml-maiden.com">[W[W(s),W](s)] = (L<sub>W(s)</sub>R + R<sup>2</sup>)(W)</formula>comparing <a href="#e200">(200)</a> and <a href="#e202">(202)</a> yields<formula xmlns="http://xml-maiden.com">(1 + c<sub>0</sub>)(L<sub>W(s)</sub>R + R<sup>2</sup>) = 2R<sup>2</sup></formula>and thus<formula xml:id="e204" xmlns="http://xml-maiden.com">(1 + c<sub>0</sub>)L<sub>W(s)</sub>R = (1 − c<sub>0</sub>)R<sup>2</sup></formula>Further let us rewrite condition <a href="#e196">(196)</a> as follows<formula xml:id="e205" xmlns="http://xml-maiden.com">W(L<sub>W(s)</sub>dJ) = c<sub>1</sub>R(W)(dJ)</formula>due to linearity of operator <math xmlns="http://xml-maiden.com">R</math> this condition can be extended to<formula xml:id="e206" xmlns="http://xml-maiden.com">R<sup>m</sup>(W)(L<sub>W(s)</sub>dJ) = c<sub>1</sub>R<sup>m + 1</sup>(W)(dJ)</formula>Now assuming that the following condition is true<formula xml:id="e207" xmlns="http://xml-maiden.com">W((L<sub>W(s)</sub>)<sup>m</sup>dJ) = c<sub>m</sub>R<sup>m</sup>(W)(dJ)</formula>let us take its Lie derivative along vector field <math xmlns="http://xml-maiden.com">W(s)</math>.We get<formula xml:id="e208" xmlns="http://xml-maiden.com">R(W)((L<sub>W(s)</sub>)<sup>m</sup>dJ) + W((L<sub>W(s)</sub>)<sup>m + 1</sup>dJ) <line/>= mc<sub>m</sub><fraction><num>1 − c<sub>0</sub></num><den>1 + c<sub>0</sub></den></fraction>R<sup>m + 1</sup>(W)(dJ) + c<sub>m</sub>R<sup>m + 1</sup>(W)(dJ)</formula>where we used properties <a href="#e199">(199)</a> and <a href="#e204">(204)</a>.Note also that <a href="#e207">(207)</a> together with linearity of operator <math xmlns="http://xml-maiden.com">R</math>imply that<formula xml:id="e209" xmlns="http://xml-maiden.com">R<sup>k</sup>W((L<sub>W(s)</sub>)<sup>m</sup>dJ) = c<sub>m</sub>R<sup>k + m</sup>(W)(dJ)</formula>and thus <a href="#e208">(208)</a> reduces to<formula xml:id="e210" xmlns="http://xml-maiden.com">W((L<sub>W(s)</sub>)<sup>m + 1</sup>dJ)= c<sub>m + 1</sub>R<sup>m + 1</sup>(W)(dJ)</formula>where <math xmlns="http://xml-maiden.com">c<sub>m + 1</sub></math> is defined by<formula xmlns="http://xml-maiden.com">(1 + c<sub>0</sub>)c<sub>m + 1</sub>= mc<sub>n</sub>(1 − c<sub>0</sub>)</formula>So we proved that if assumtion <a href="#e207">(207)</a> is valid for <math xmlns="http://xml-maiden.com">m</math>then it is also valid for <math xmlns="http://xml-maiden.com">m + 1</math>, we also know that for <math xmlns="http://xml-maiden.com">m = 1</math> itmatches <a href="#e205">(205)</a> and thus by induction we proved that condition<a href="#e207">(207)</a> is valid for arbitrary <math xmlns="http://xml-maiden.com">m</math> while <math xmlns="http://xml-maiden.com">c<sub>n</sub></math>can be determined by<formula xmlns="http://xml-maiden.com">c<sub>m</sub>(1 + c<sub>0</sub>)<sup>m − 1</sup> = c<sub>0</sub>(m − 1)!(1 − c<sub>0</sub>)<sup>m − 1</sup></formula>Now using <a href="#e207">(207)</a> and <a href="#e209">(209)</a>it is easy to show that functions <math xmlns="http://xml-maiden.com">(L<sub>W(s)</sub>)<sup>m</sup>J</math> are in involution.Indeed<formula xmlns="http://xml-maiden.com">{(L<sub>W(s)</sub>)<sup>m</sup>J, (L<sub>W(s)</sub>)<sup>k</sup>J} =W(d(L<sub>W(s)</sub>)<sup>m</sup>J ∧ d(L<sub>W(s)</sub>)<sup>k</sup>J) <line/>= W((L<sub>W(s)</sub>)<sup>m</sup>dJ ∧ (L<sub>W(s)</sub>)<sup>k</sup>dJ) =c<sub>m</sub>c<sub>k</sub>W(dJ ∧ dJ) = 0</formula>So we have proved that the functions <a href="#e197">(197)</a> are in involution.</div><div class="paragraph">Further we will use this theorem to prove involutivity of family of conservation laws constructed using non-Noether symmetry of Toda chain.</div><h2>Toda Model</h2><div class="paragraph">To illustrate features of non-Noether symmetries we oftenrefer to two and three particle non-periodic Toda systems.However it turns out that non-Noether symmetries are present ingeneric n-particle non-periodic Toda chains as well, moreover they preservebasic features of symmetries <a href="#e53">(53)</a>, <a href="#e61">(61)</a>.In case of n-particle Toda model symmetry yields <math xmlns="http://xml-maiden.com">n</math>functionally independent conservation laws in involution,gives rise to bi-Hamiltonian structure of Toda hierarchy,reproduces Lax pair of Toda system, endows phase space withFrölicher-Nijenhuis operator and leads to invariantbidifferential calculus on algebra of differential forms over phase spaceof Toda system.</div><div class="paragraph">First of all let us remind that Toda model is<math xmlns="http://xml-maiden.com">2n</math> dimensional Hamiltonian system that describes the motionof <math xmlns="http://xml-maiden.com">n</math> particles on the line governed by the exponential interaction.Equations of motion of the non periodic n-particle Toda model are<formula xml:id="e214" xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>q<sub>s</sub> = p<sub>s</sub><line/><fraction><num>d</num><den>dt</den></fraction>p<sub>s</sub> = ε(s − 1)e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup> −ε(n − s)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup></formula>(<math xmlns="http://xml-maiden.com">ε(k) = − ε(− k) = 1</math> for any natural<math xmlns="http://xml-maiden.com">k</math> and <math xmlns="http://xml-maiden.com">ε(0) = 0</math>) and can be rewritten in Hamiltonian form<a href="#e24">(24)</a> with canonical Poisson bracket defined by Poisson bivector<formula xmlns="http://xml-maiden.com">W = <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> <fraction><num>∂</num><den>∂p<sub>s</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂q<sub>s</sub></den></fraction></formula>and Hamiltonian equal to<formula xmlns="http://xml-maiden.com">h = ½<under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under>p<sub>s</sub><sup>2</sup> +<under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under>e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup></formula>Note that in two and three particle case we have used slightly different notations<formula xmlns="http://xml-maiden.com">z<sub>s</sub> = p<sub>s</sub><line/>z<sub>n + s</sub> = q<sub>s</sub>        s = 1, 2, (3); n = 2(3)</formula>for local coordinates.The group of transformations <math xmlns="http://xml-maiden.com">g<sub>z</sub></math> generated by the vector field<math xmlns="http://xml-maiden.com">E</math> will be symmetry of Toda chain if for each<math xmlns="http://xml-maiden.com">p<sub>s</sub>, q<sub>s</sub></math> satisfying Toda equations<a href="#e214">(214)</a><math xmlns="http://xml-maiden.com">g<sub>z</sub>(p<sub>s</sub>), g<sub>z</sub>(q<sub>s</sub>)</math>also satisfy it.Substituting infinitesimal transformations<formula xmlns="http://xml-maiden.com">g<sub>z</sub>(p<sub>s</sub>) = p<sub>s</sub> + zE(p<sub>s</sub>) + O(z<sup>2</sup>)<line/>g<sub>z</sub>(p<sub>s</sub>) = q<sub>s</sub> + zE(q<sub>s</sub>) + O(z<sup>2</sup>)</formula>into <a href="#e214">(214)</a> and grouping first order terms gives rise to theconditions<formula xml:id="e219" xmlns="http://xml-maiden.com"><fraction><num>d</num><den>dt</den></fraction>E(q<sub>s</sub>) = E(p<sub>s</sub>)<line/><fraction><num>d</num><den>dt</den></fraction>E(p<sub>s</sub>) = ε(s − 1)e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup>(E(q<sub>s − 1</sub>) − E(q<sub>s</sub>)) − ε(n − s)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup>(E(q<sub>s</sub>) − E(q<sub>s + 1</sub>))</formula>One can verify that the vector field defined by<formula xml:id="e220" xmlns="http://xml-maiden.com">E(p<sub>s</sub>) = ½p<sub>s</sub><sup>2</sup> +ε(s − 1)(n − s + 2)e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup> −ε(n − s)(n − s) e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>+ <fraction><num>t</num><den>2</den></fraction>(ε(s − 1)(p<sub>s − 1</sub> + p<sub>s</sub>)e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup> −ε(n − s)(p<sub>s</sub> + p<sub>s + 1</sub>)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup>)<line/>E(q<sub>s</sub>) = (n − s + 1)p<sub>s</sub> −½<under><over><sat>s − 1</sat><ope>∑</ope></over><sat>k = 1</sat></under> p<sub>k</sub>+ ½<under><over><sat>n</sat><ope>∑</ope></over><sat>k = s + 1</sat></under> p<sub>k</sub><line/>+ <fraction><num>t</num><den>2</den></fraction>(p<sub>s</sub><sup>2</sup> +ε(s − 1)e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup> +ε(n − s)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup>)</formula>satisfies <a href="#e31">(31)</a> and generates symmetry of Toda chain. It appears that this symmetry is non-Noether since it does notpreserve Poisson bracket structure <math xmlns="http://xml-maiden.com">[E , W] ≠ 0</math>and additionally one can check that Yang-Baxter equation<math xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</math> is satisfied.This symmetry may play important role inanalysis of Toda model. First let us note that calculating <math xmlns="http://xml-maiden.com">L<sub>E</sub>W</math>leads to the following Poisson bivector field<formula xmlns="http://xml-maiden.com">Ŵ = [E , W] =<under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> p<sub>s</sub><fraction><num>∂</num><den>∂p<sub>s</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂q<sub>s</sub></den></fraction>+ <under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup> <fraction><num>∂</num><den>∂p<sub>s</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂q<sub>s + 1</sub></den></fraction><line/>+ <under><ope>∑</ope><sat>r &gt; s</sat></under> <fraction><num>∂</num><den>∂q<sub>s</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂q<sub>r</sub></den></fraction></formula>and together <math xmlns="http://xml-maiden.com">W</math> and <math xmlns="http://xml-maiden.com">L<sub>E</sub>W</math> give rise tobi-Hamiltonian structure of Toda model (compare with <a href="#r30">[30]</a>).Thus bi-Hamiltonian realization of Toda chain can be considered as manifestationof hidden symmetry.In terms of bivector fields these bi-Hamiltonian system is formed byThe conservation laws <a href="#e45">(45)</a> associated with the symmetry reproduce well knownset of conservation laws of Toda chain.<formula xml:id="e222" xmlns="http://xml-maiden.com">I<sup>(1)</sup> = C<sup>(1)</sup> = <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under>p<sub>s</sub><line/>I<sup>(2)</sup> = (C<sup>(1)</sup>)<sup>2</sup> − 2C<sup>(2)</sup> =<under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> p<sub>s</sub><sup>2</sup> + 2<under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under>e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>I<sup>(3)</sup> = C<sup>(1)</sup>)<sup>3</sup> − 3C<sup>(1)</sup>C<sup>(2)</sup>+ 3C<sup>(3)</sup> = <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> p<sub>s</sub><sup>3</sup> +3<under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> (p<sub>s</sub> + p<sub>s + 1</sub>) e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>I<sup>(4)</sup> = C<sup>(1)</sup>)<sup>4</sup> − 4(C<sup>(1)</sup>)<sup>2</sup>C<sup>(2)</sup> +2(C<sup>(2)</sup>)<sup>2</sup> + 4C<sup>(1)</sup>C<sup>(3)</sup> − 4C<sup>(4)</sup><line/>= <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> p<sub>s</sub><sup>4</sup> + 4<under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under>(p<sub>s</sub><sup>2</sup> + 2p<sub>s</sub>p<sub>s + 1</sub> + p<sub>s + 1</sub><sup>2</sup>)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>+ 2<under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>2(q<sub>s</sub> − q<sub>s + 1</sub>)</sup> +4<under><over><sat>n − 2</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>q<sub>s</sub> − q<sub>s + 2</sub></sup> <line/>I<sup>(m)</sup> = (− 1)<sup>m + 1</sup>mC<sup>(m)</sup> +<under><over><sat>m − 1</sat><ope>∑</ope></over><sat>k = 1</sat></under>(− 1)<sup>k + 1</sup>I<sup>(m − k)</sup>C<sup>(k)</sup></formula>The condition <math xmlns="http://xml-maiden.com">[[E[E , W]]W] = 0</math> satisfied by generator of thesymmetry <math xmlns="http://xml-maiden.com">E</math> ensures that the conservation laws are in involutioni. e. <math xmlns="http://xml-maiden.com">{C<sup>(k)</sup> , C<sup>(m)</sup>} = 0</math>.Thus the conservation laws as well as the bi-Hamiltonian structureof the non periodic Toda chain appear to be associated with non-Noether symmetry.</div><div class="paragraph">Using formula <a href="#e88">(88)</a> one can calculate Lax pairassociated with symmetry <a href="#e220">(220)</a>.Lax matrix calculated in this way has the following non-zero entries(note that in case of <math xmlns="http://xml-maiden.com">n = 2</math> and <math xmlns="http://xml-maiden.com">n = 3</math> this formula yields matrices<a href="#e102">(102)</a>-<a href="#e105">(105)</a>)<formula xml:id="e223" xmlns="http://xml-maiden.com">L<sub>k, k</sub> = L<sub>n + k, n + k</sub> = p<sub>k</sub><line/>L<sub>n + k, k + 1</sub> = − L<sub>n + k + 1, k</sub> =ε(n − k)e<sup>q<sub>k</sub> − q<sub>k + 1</sub></sup><line/>L<sub>k, n + m</sub> = ε(m − k)<line/>m, k = 1, 2, ... , n</formula>while non-zero entries of <math xmlns="http://xml-maiden.com">P</math> matrix involved in Lax pair are<formula xml:id="e224" xmlns="http://xml-maiden.com">P<sub>k, n + k</sub> = 1<line/>P<sub>n + k, k</sub> = − ε(k − 1)e<sup>q<sub>k − 1</sub> − q<sub>k</sub></sup>− ε(n − k)e<sup>q<sub>k</sub> − q<sub>k + 1</sub></sup><line/>P<sub>n + k, k + 1</sub> = ε(n − k)e<sup>q<sub>k</sub> − q<sub>k + 1</sub></sup><line/>P<sub>n + k, k − 1</sub> = ε(k − 1)e<sup>q<sub>k − 1</sub> − q<sub>k</sub></sup><line/>k = 1, 2, ... , n</formula>This Lax pair constructed from generator of non-Noether symmetryexactly reproduces known Lax pair of Toda chain.</div><div class="paragraph">Like two and three particle Toda chain, n-particle Toda model also admitsinvariant bidifferential calculus on algebra of differential forms over the phase space.This bidifferential calculus can be constructed using non-Noether symmetry (see <a href="#e152">(152)</a>),it consists out of two differential operators <math xmlns="http://xml-maiden.com">d, đ</math>where <math xmlns="http://xml-maiden.com">d</math> is ordinary exterior derivative while <math xmlns="http://xml-maiden.com">đ</math>can be defined by<formula xml:id="e225" xmlns="http://xml-maiden.com">đq<sub>s</sub> = p<sub>s</sub>dq<sub>s</sub> + <under><ope>∑</ope><sat>r &gt; s</sat></under>dp<sub>r</sub> − <under><ope>∑</ope><sat>s &gt; r</sat></under>dp<sub>r</sub><line/>đp<sub>s</sub> = p<sub>s</sub>dp<sub>s</sub> − e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup>dq<sub>s + 1</sub>+ e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup>dq<sub>s</sub></formula>and is extended to whole De Rham complex by linearity, derivation property andcompatibility property <math xmlns="http://xml-maiden.com">dđ + đd = 0</math>.By direct calculations one can verify that calculus constructed in this wayis consistent and satisfies <math xmlns="http://xml-maiden.com">đ<sup>2</sup> = 0</math> property.One can also check that conservation laws <a href="#e222">(222)</a> form Lenard scheme<formula xmlns="http://xml-maiden.com">(k + 1)đI<sup>(k)</sup> = kdI<sup>(k + 1)</sup></formula></div><div class="paragraph">Further let us focus on Frölicher-Nijenhuis geometry. Using formula <a href="#e173">(173)</a>one can construct invariant Frölicher-Nijenhuis operator, out of generator of non-Noethersymmetry of Toda chain. Operator constructed in this way has the form<formula xml:id="e227" xmlns="http://xml-maiden.com">Ŕ<sub>E</sub> = <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under>p<sub>s</sub>(dp<sub>s</sub> ⊗ <fraction><num>∂</num><den>∂q<sub>s</sub></den></fraction> + dq<sub>s</sub> ⊗ <fraction><num>∂</num><den>∂p<sub>s</sub></den></fraction>)<line/> − <under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup>dq<sub>s + 1</sub> ⊗ <fraction><num>∂</num><den>∂p<sub>s</sub></den></fraction>+ <under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>q<sub>s − 1</sub> − q<sub>s</sub></sup>dq<sub>s</sub> ⊗ <fraction><num>∂</num><den>∂p<sub>s</sub></den></fraction><line/>− <under><ope>∑</ope><sat>s &gt; r</sat></under> (dp<sub>s</sub> ⊗ <fraction><num>∂</num><den>∂q<sub>r</sub></den></fraction> − dp<sub>r</sub> ⊗ <fraction><num>∂</num><den>∂q<sub>s</sub></den></fraction>)</formula>One can check that Frölicher-Nijenhuis torsion of this operator vanishes andit plays role of recursion operator for n-particle Toda chain in sense that conservation laws<math xmlns="http://xml-maiden.com">I<sup>(k)</sup></math> satisfy recursion relation<formula xml:id="e228" xmlns="http://xml-maiden.com">(k + 1)R<sub>E</sub>(dI<sup>(k)</sup>) = kdI<sup>(k + 1)</sup></formula>Thus non-Noether symmetry of Toda chain not only leads ton functionally independent conservation laws in involution, but alsoessentially enriches phase space geometry by endowing it withinvariant Frölicher-Nijenhuis operator, bi-Hamiltonian system,bicomplex structure and Lax pair.</div><div class="paragraph">Finally, in order to outline possible applications of Theorem 8 let us studyaction of non-Noether symmetry <a href="#e220">(220)</a> on conserved quantitiesof Toda chain.  Vector field <math xmlns="http://xml-maiden.com">E</math> defined by <a href="#e220">(220)</a> generatesone-parameter group of transformations <a href="#e28">(28)</a> that maps arbitraryconserved quantity <math xmlns="http://xml-maiden.com">J</math> to<formula xmlns="http://xml-maiden.com">J(z) = J + zJ<sup>(1)</sup> + <fraction><num>z<sup>2</sup></num><den>2!</den></fraction>J<sup>(2)</sup> +<fraction><num>z<sup>3</sup></num><den>3!</den></fraction>J<sup>(3)</sup> + ⋯</formula>where<formula xmlns="http://xml-maiden.com">J<sup>(m)</sup> = (L<sub>E</sub>)<sup>m</sup>J</formula>In particular let us focus on family of conserved quantities obtained by action of<math xmlns="http://xml-maiden.com">g<sub>a</sub> = e<sup>aL<sub>E</sub></sup></math> on total momenta of Toda chain<formula xml:id="e231" xmlns="http://xml-maiden.com">J = <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> p<sub>s</sub></formula>By direct calculations one can check that family <math xmlns="http://xml-maiden.com">J(z)</math>, that forms orbitof non-Noether symmetry generated by <math xmlns="http://xml-maiden.com">E</math>, reproduces entire involutivefamily of integrals of motion <a href="#e222">(222)</a>. Namely<formula xml:id="e232" xmlns="http://xml-maiden.com">J<sup>(1)</sup> = L<sub>E</sub>J = ½ <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under> p<sub>s</sub><sup>2</sup> + <under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under>e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> = (L<sub>E</sub>)<sup>2</sup>J =<fraction><num>1</num><den>2</den></fraction> <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under>p<sub>s</sub><sup>3</sup> +<fraction><num>3</num><den>2</den></fraction><under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> (p<sub>s</sub> + p<sub>s + 1</sub>)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>J<sup>(3)</sup> = L<sub>E</sub>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>3</sup>J =¾ <under><over><sat>n</sat><ope>∑</ope></over><sat>s = 1</sat></under>p<sub>s</sub><sup>4</sup> +3<under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under>(p<sub>s</sub><sup>2</sup> + 2p<sub>s</sub>p<sub>s + 1</sub> +p<sub>s + 1</sub><sup>2</sup>)e<sup>q<sub>s</sub> − q<sub>s + 1</sub></sup><line/>+ <fraction><num>3</num><den>2</den></fraction> <under><over><sat>n − 1</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>2(q<sub>s</sub> − q<sub>s + 1</sub>)</sup> +3<under><over><sat>n − 2</sat><ope>∑</ope></over><sat>s = 1</sat></under> e<sup>q<sub>s</sub> − q<sub>s + 2</sub></sup><line/>J<sup>(m)</sup> = L<sub>E</sub>J<sup>(m − 1)</sup> = (L<sub>E</sub>)<sup>m</sup>J</formula></div><div class="paragraph">Involutivity of this set of conservation laws can be verified using Theorem 8.In particular one can notice that differential 1-form <math xmlns="http://xml-maiden.com">s</math> defined by<formula xmlns="http://xml-maiden.com">E = W(s)</formula>(where <math xmlns="http://xml-maiden.com">E</math> is generator of non-Noether symmetry <a href="#e220">(220)</a>)satisfies condition<formula xmlns="http://xml-maiden.com">[W[W(s),W](s)] = 3[W(s)[W(s) ,W]]</formula>while conservation law <math xmlns="http://xml-maiden.com">J</math> defined by <a href="#e231">(231)</a>has property<formula xmlns="http://xml-maiden.com">W(L<sub>W(s)</sub>dJ) = − [W(s),W](dJ)</formula>and thus according to Theorem 8 conservation laws <a href="#e232">(232)</a>are in involution.</div><h2>Korteweg-de Vries equation</h2><div class="paragraph">Toda model provided good example of finite dimensional integrable Hamiltonian systemthat possesses non-Noether symmetry. However there are manyinfinite dimensional integrable Hamiltonian systems and in this case inorder to ensure integrability one should constructinfinite number of conservation laws. Fortunately in several integrable modelsthis task can be effectively simplified by identifying appropriate non-Noether symmetry.First let us consider well known infinite dimensional integrable Hamiltonian system –Korteweg-de Vries equation (KdV). The KdV equation has the following form<formula xml:id="e236" xmlns="http://xml-maiden.com">u<sub>t</sub> + u<sub>xxx</sub> + uu<sub>x</sub> = 0</formula>(here <math xmlns="http://xml-maiden.com">u</math> is smooth function of <math xmlns="http://xml-maiden.com">(t, x) ∈ R<sup>2</sup></math>).The generators of symmetries of KdV should satisfy condition<formula xml:id="e237" xmlns="http://xml-maiden.com">E(u)<sub>t</sub> + E(u)<sub>xxx</sub> +u<sub>x</sub>E(u) + uE(u)<sub>x</sub> = 0</formula>which is obtained by substituting infinitesimal transformation<math xmlns="http://xml-maiden.com">u  →  u + zE(u) + O(z<sup>2</sup>)</math> into KdV equation and grouping first order terms.</div><div class="paragraph">Later we will focus on the symmetry generated by the following vector field<formula xml:id="e238" xmlns="http://xml-maiden.com">E(u) = 2u<sub>xx</sub> + <fraction><num>2</num><den>3</den></fraction>u<sup>2</sup> + <fraction><num>1</num><den>6</den></fraction>u<sub>x</sub>v +<fraction><num>x</num><den>2</den></fraction>(u<sub>xxx</sub> + uu<sub>x</sub>) − <line/><fraction><num>t</num><den>4</den></fraction>(6u<sub>xxxxx</sub> + 20u<sub>x</sub>u<sub>xx</sub> +10 uu<sub>xxx</sub> + 5u<sup>2</sup>u<sub>x</sub>)</formula>(here <math xmlns="http://xml-maiden.com">v</math> is defined by <math xmlns="http://xml-maiden.com">v<sub>x</sub> = u</math>).</div><div class="paragraph">If <math xmlns="http://xml-maiden.com">u</math> is subjected to zero  boundary conditions <math xmlns="http://xml-maiden.com">u(t, − ∞) = u(t, + ∞) = 0</math>then KdV equation can be rewritten in Hamiltonian form <formula xml:id="e239" xmlns="http://xml-maiden.com">u<sub>t</sub> = {h , u}</formula>with Poisson bivector field equal to<formula xml:id="e240" xmlns="http://xml-maiden.com">W = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>dx <fraction><num>δ</num><den>δu</den></fraction>∧ {<fraction><num>δ</num><den>δu</den></fraction>}<sub>x</sub></formula>and Hamiltonian defined by<formula xml:id="e241" xmlns="http://xml-maiden.com">h = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sub>x</sub><sup>2</sup> − <fraction><num>u<sup>3</sup></num><den>3</den></fraction>) dx</formula>By taking Lie derivative of thesymplectic form along the generator of the symmetry one getssecond Poisson bivector <formula xml:id="e242" xmlns="http://xml-maiden.com">[E , W] = W = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>dx ({<fraction><num>δ</num><den>δu</den></fraction>}<sub>xx</sub> ∧ {<fraction><num>δ</num><den>δu</den></fraction>}<sub>x</sub>+ <fraction><num>2</num><den>3</den></fraction>u<fraction><num>δ</num><den>δu</den></fraction> ∧ {<fraction><num>δ</num><den>δu</den></fraction>}<sub>x</sub>)</formula>involved in bi-Hamiltonian structure of KdV hierarchy andproposed by Magri <a href="#r58">[58]</a>.</div><div class="paragraph">Now let us show how non-Noether symmetry can be used to construct conservation lawsof KdV hierarchy. By integrating KdV it is easy to show that<formula xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>u dx</formula>is conserved quantity. At the same time Lie derivative of any conservedquantity along generator of symmetry is conserved as well,while taking Lie derivative of <math xmlns="http://xml-maiden.com">J<sup>(0)</sup></math> along <math xmlns="http://xml-maiden.com">E</math> gives rise toinfinite sequence of conservation laws <math xmlns="http://xml-maiden.com">J<sup>(m)</sup> = (L<sub>E</sub>)<sup>m</sup>J<sup>(0)</sup></math>that reproduce well known conservation laws of KdV equation<formula xml:id="e244" xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>u dx<line/>J<sup>(1)</sup> = L<sub>E</sub>J<sup>(0)</sup> =¼<under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>u<sup>2</sup> dx <line/>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>2</sup>J<sup>(0)</sup> =<fraction><num>5</num><den>8</den></fraction><under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>(<fraction><num>u<sup>3</sup></num><den>3</den></fraction> − u<sub>x</sub><sup>2</sup>) dx <line/>J<sup>(3)</sup> = (L<sub>E</sub>)<sup>3</sup>J<sup>(0)</sup> <line/>= <fraction><num>35</num><den>16</den></fraction><under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (<fraction><num>5</num><den>36</den></fraction>u<sup>4</sup> −<fraction><num>5</num><den>3</den></fraction>uu<sub>x</sub><sup>2</sup> + u<sub>xx</sub><sup>2</sup>) dx<line/>J<sup>(m)</sup> = (L<sub>E</sub>)<sup>m</sup>J<sup>(0)</sup></formula>Thus the conservation laws and bi-Hamiltonian structures of KdV hierarchy are related to the non-Noether symmetry of KdV equation.</div><h2>Nonlinear water wave equations</h2><div class="paragraph">Among nonlinear partial differential equations that describe propagation of waves in shallow waterthere are many remarkable integrable systems. We have already discussed case of KdV equation, that possess non-Noether symmetries leading to the infinite sequence of conservation lawsand bi-Hamiltonian realization of these equations,now let us consider other important water wave systems.It is reasonable to start with dispersive water wave system <a href="#r73">[73]</a>,<a href="#r74">[74]</a>,since many other models can be obtained from it by reduction.Evolution of dispersive water wave system is governed bythe following set of equations<formula xml:id="e245" xmlns="http://xml-maiden.com">u<sub>t</sub> = u<sub>x</sub>w + uw<sub>x</sub><line/>v<sub>t</sub> = uu<sub>x</sub> − v<sub>xx</sub> + 2v<sub>x</sub>w + 2vw<sub>x</sub><line/>w<sub>t</sub> = w<sub>xx</sub> − 2v<sub>x</sub> + 2ww<sub>x</sub></formula>Each symmetry of this system must satisfy linear equation<formula xmlns="http://xml-maiden.com">E(u)<sub>t</sub> = (wE(u))<sub>x</sub> + (uE(w))<sub>x</sub><line/>E(v)<sub>t</sub> = (uE(u))<sub>x</sub> − E(v)<sub>xx</sub> + 2(wE(v))<sub>x</sub> + 2(vE(w))<sub>x</sub><line/>E(w)<sub>t</sub> = E(w)<sub>xx</sub> − 2E(v)<sub>x</sub> + 2(wE(w))<sub>x</sub></formula>obtained by substituting infinitesimal transformations<formula xmlns="http://xml-maiden.com">u  → u + zE(u) + O(z<sup>2</sup>)<line/>v  →  v + zE(v) + O(z<sup>2</sup>)<line/>w  →  w + zE(w) + O(z<sup>2</sup>)</formula>into equations of motion <a href="#e245">(245)</a> and grouping first order(in <math xmlns="http://xml-maiden.com">a</math>) terms. One of the solutions of this equation yieldsthe following symmetry of dispersive water wave system<formula xml:id="e248" xmlns="http://xml-maiden.com">E(u) = uw + x(uw)<sub>x</sub> + 2t(uw<sup>2</sup> − 2uv + uw<sub>x</sub>)<sub>x</sub><line/>E(v) = <fraction><num>3</num><den>2</den></fraction>u<sup>2</sup> + 4vw − 3v<sub>x</sub> + x(uu<sub>x</sub> + 2(vw)<sub>x</sub> − v<sub>xx</sub>)<line/>+ 2t(u<sup>2</sup>w − uu<sub>x</sub> − 3v<sup>2</sup> + 3vw<sup>2</sup> − 3v<sub>x</sub>w + v<sub>xx</sub>)<sub>x</sub><line/>E(w) = w<sup>2</sup> + 2w<sub>x</sub> − 4v + x(2ww<sub>x</sub> + w<sub>xx</sub> − 2v<sub>x</sub>)<line/>− 2t(u<sup>2</sup> + 6vw − w<sup>3</sup> − 3ww<sub>x</sub> − w<sub>xx</sub>)<sub>x</sub></formula>and it is remarkable that this symmetry is local in sense that <math xmlns="http://xml-maiden.com">E(u)</math> in point<math xmlns="http://xml-maiden.com">x</math> depends only on <math xmlns="http://xml-maiden.com">u</math> and its derivatives evaluated in the same point,(this is not the case in KdV where symmetry is non localdue to presence of non local field <math xmlns="http://xml-maiden.com">v</math> defined by <math xmlns="http://xml-maiden.com">v<sub>x</sub> = u</math>).</div><div class="paragraph">Before we proceed let us note that dispersive water wave system is actually infinite dimensionalHamiltonian dynamical system. Assuming that <math xmlns="http://xml-maiden.com">u, v</math> and <math xmlns="http://xml-maiden.com">w</math> fieldsare subjected to zero boundary conditions<formula xmlns="http://xml-maiden.com">u(± ∞) = v(± ∞) = w(± ∞) = 0</formula>it is easy to verify that equations <a href="#e245">(245)</a> can be represented in Hamiltonian form<formula xmlns="http://xml-maiden.com">u<sub>t</sub> = {h , u}<line/>v<sub>t</sub> = {h , v}<line/>w<sub>t</sub> = {h , w}</formula>with Hamiltonian equal to<formula xml:id="e251" xmlns="http://xml-maiden.com">h = − ¼ <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup>w + 2vw<sup>2</sup> − 2v<sub>x</sub>w − 2v<sup>2</sup>)dx</formula>and Poisson bracket defined by the following Poisson bivector field<formula xml:id="e252" xmlns="http://xml-maiden.com">W = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> {½ <fraction><num>δ</num><den>δu</den></fraction> ∧ {<fraction><num>δ</num><den>δu</den></fraction>}<sub>x</sub> +<fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx</formula>Now using our symmetry that appears to be non-Noether, one can calculate second Poissonbivector field involved in the bi-Hamiltonian realization of dispersive water wave system<formula xml:id="e253" xmlns="http://xml-maiden.com">Ŵ = [E , W] = − 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>{u <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δu</den></fraction>}<sub>x</sub>+ v <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub><line/>+ {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>+ w <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>+ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub> ∧ <fraction><num>δ</num><den>δw</den></fraction>} dx</formula>Note that <math xmlns="http://xml-maiden.com">Ŵ</math> give rise to the second Hamiltonian realization ofthe model<formula xmlns="http://xml-maiden.com">u<sub>t</sub> = {h<sup>∗</sup> , u}<sub>∗</sub><line/>v<sub>t</sub> = {h<sup>∗</sup> , v}<sub>∗</sub><line/>w<sub>t</sub> = {h<sup>∗</sup> , w}<sub>∗</sub></formula>where<formula xmlns="http://xml-maiden.com">h<sup>∗</sup> = − ¼ <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup> + 2vw)dx</formula>and <math xmlns="http://xml-maiden.com">{ , }<sub>∗</sub></math> is Poisson bracket defined bybivector field <math xmlns="http://xml-maiden.com">Ŵ</math>.</div><div class="paragraph">Now let us pay attention to conservation laws. By integrating third equationof dispersive water wave system <a href="#e245">(245)</a> it is easy to show that<formula xmlns="http://xml-maiden.com">J<sup>(0)</sup> =<under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>wdx</formula>is conservation law. Using non-Noether symmetryone can construct other conservation laws by taking Lie derivativeof <math xmlns="http://xml-maiden.com">J<sup>(0)</sup></math> along the generator of symmetry and in this wayentire infinite sequence of conservation laws of dispersive water wave systemcan be reproduced<formula xml:id="e257" xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> wdx<line/>J<sup>(1)</sup> = L<sub>E</sub>J<sup>(0)</sup> = − 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vdx <line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> = (L<sub>E</sub>)<sup>2</sup>J<sup>(0)</sup> =− 2<under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup> + 2vw)dx<line/>J<sup>(3)</sup> = L<sub>E</sub>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>3</sup>J<sup>(0)</sup> =− 6<under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup>w + 2vw<sup>2</sup> − 2v<sub>x</sub>w − 2v<sup>2</sup>)dx<line/>J<sup>(4)</sup> = L<sub>E</sub>J<sup>(3)</sup> = (L<sub>E</sub>)<sup>4</sup>J<sup>(0)</sup><line/>= − 24 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup>w<sup>2</sup> + u<sup>2</sup>w<sub>x</sub> − 2u<sup>2</sup>v − 6v<sup>2</sup>w +2vw<sup>3</sup> − 3v<sub>x</sub>w<sup>2</sup> − 2v<sub>x</sub>w<sub>x</sub>)dx<line/>J<sup>(n)</sup> = L<sub>E</sub>J<sup>(n − 1)</sup> = (L<sub>E</sub>)<sup>n</sup>J<sup>(0)</sup></formula>Thus conservation laws and bi-Hamiltonian structure of dispersive waterwave system can be constructed by means of non-Noether symmetry.</div><div class="paragraph">Note that symmetry <a href="#e248">(248)</a> can be used in many otherpartial differential equations that can be obtained by reduction from dispersivewater wave system. In particular one can use it in dispersiveless water wave system,Broer-Kaup system, dispersiveless long wave system, Burger's equation etc.In case of dispersiveless water waves system<formula xml:id="e258" xmlns="http://xml-maiden.com">u<sub>t</sub> = u<sub>x</sub>w + uw<sub>x</sub><line/>v<sub>t</sub> = uu<sub>x</sub> + 2v<sub>x</sub>w + 2vw<sub>x</sub><line/>w<sub>t</sub> = − 2v<sub>x</sub> + 2ww<sub>x</sub></formula>symmetry <a href="#e248">(248)</a> is reduced to<formula xml:id="e259" xmlns="http://xml-maiden.com">E(u) = uw + x(uw)<sub>x</sub> + 2t(uw<sup>2</sup> − 2uv)<sub>x</sub><line/>E(v) = <fraction><num>3</num><den>2</den></fraction>u<sup>2</sup> + 4vw + x(uu<sub>x</sub> + 2(vw)<sub>x</sub>)+ 2t(u<sup>2</sup>w − 3v<sup>2</sup> + 3vw<sup>2</sup>)<sub>x</sub><line/>E(w) = w<sup>2</sup> − 4v + x(2ww<sub>x</sub> − 2v<sub>x</sub>) − 2t(u<sup>2</sup> + 6vw − w<sup>3</sup>)<sub>x</sub></formula>and corresponding conservation laws <a href="#e257">(257)</a> reduce to<formula xml:id="e260" xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> wdx<line/>J<sup>(1)</sup> = L<sub>E</sub>J<sup>(0)</sup> =− 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vdx<line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> = (L<sub>E</sub>)<sup>2</sup>J<sup>(0)</sup> =− 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup> + 2vw)dx<line/>J<sup>(3)</sup> = L<sub>E</sub>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>3</sup>J<sup>(0)</sup> =− 6 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup>w + 2vw<sup>2</sup> − 2v<sup>2</sup>)dx<line/>J<sup>(4)</sup> = L<sub>E</sub>J<sup>(3)</sup> = (L<sub>E</sub>)<sup>4</sup>J<sup>(0)</sup> =− 24 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (u<sup>2</sup>w<sup>2</sup> − 2u<sup>2</sup>v − 6v<sup>2</sup>w + 2vw<sup>3</sup>)dx<line/>J<sup>(n)</sup> = L<sub>E</sub>J<sup>(n − 1)</sup> = (L<sub>E</sub>)<sup>n</sup>J<sup>(0)</sup></formula></div><div class="paragraph">Another important integrable model that can be obtained from dispersive water wave systemis Broer-Kaup system <a href="#r73">[73]</a>,<a href="#r74">[74]</a><formula xml:id="e261" xmlns="http://xml-maiden.com">v<sub>t</sub> = ½ v<sub>xx</sub> + v<sub>x</sub>w + vw<sub>x</sub><line/>w<sub>t</sub> = − ½ w<sub>xx</sub> + v<sub>x</sub> + ww<sub>x</sub></formula>One can check that symmetry <a href="#e248">(248)</a> of dispersive water wave system,after reduction, reproduces non-Noether symmetry of Broer-Kaup model<formula xml:id="e262" xmlns="http://xml-maiden.com">E(v) = 4vw + 3v<sub>x</sub> + x(2(vw)<sub>x</sub> + v<sub>xx</sub>)<line/>+ t(3v<sup>2</sup> + 3vw<sup>2</sup> + 3v<sub>x</sub>w + v<sub>xx</sub>)<sub>x</sub><line/>E(w) = w<sup>2</sup> − 2w<sub>x</sub> + 4v + x(2ww<sub>x</sub> − w<sub>xx</sub> + 2v<sub>x</sub>)<line/>+ t(6vw + w<sup>3</sup> − 3ww<sub>x</sub> + w<sub>xx</sub>)<sub>x</sub></formula>and gives rise to the infinite sequence of conservation laws of Broer-Kaup hierarchy<formula xml:id="e263" xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> wdx<line/>J<sup>(1)</sup> = L<sub>E</sub>J<sup>(0)</sup> = 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vdx<line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> = (L<sub>E</sub>)<sup>2</sup>J<sup>(0)</sup> =4 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vwdx<line/>J<sup>(3)</sup> = L<sub>E</sub>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>3</sup>J<sup>(0)</sup> =12 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (vw<sup>2</sup> + v<sub>x</sub>w + v<sup>2</sup>)dx<line/>J<sup>(4)</sup> = L<sub>E</sub>J<sup>(3)</sup> = (L<sub>E</sub>)<sup>4</sup>J<sup>(0)</sup> =24 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (6v<sup>2</sup>w + 2vw<sup>3</sup> + 3v<sub>x</sub>w<sup>2</sup> − 2v<sub>x</sub>w<sub>x</sub>)dx<line/>J<sup>(n)</sup> = L<sub>E</sub>J<sup>(n − 1)</sup> = (L<sub>E</sub>)<sup>n</sup>J<sup>(0)</sup></formula></div><div class="paragraph">And exactly like in the dispersive water wave system one can rewrite equations of motion<a href="#e261">(261)</a> in Hamiltonian form<formula xmlns="http://xml-maiden.com">v<sub>t</sub> = {h , v}<line/>w<sub>t</sub> = {h , w}</formula>where Hamiltonian is<formula xmlns="http://xml-maiden.com">h = ½ <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (vw<sup>2</sup> + v<sub>x</sub>w + v<sup>2</sup>)dx</formula>while Poisson bracket is defined by the Poisson bivector field<formula xml:id="e266" xmlns="http://xml-maiden.com">W = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> {<fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx</formula>And again, using symmetry <a href="#e262">(262)</a> one can recover second Poissonbivector field involved in the bi-Hamiltonian realization of Broer-Kaup systemby taking Lie derivative of <a href="#e266">(266)</a><formula xml:id="e267" xmlns="http://xml-maiden.com">Ŵ = [E , W] = − 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> {v <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub><line/>− {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub> + w <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>+ <fraction><num>δ</num><den>δw</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx</formula>This bivector field give rise to the second Hamiltonian realization ofthe Broer-Kaup system<formula xmlns="http://xml-maiden.com">v<sub>t</sub> = {h<sup>∗</sup> , v}<sub>∗</sub><line/>w<sub>t</sub> = {h<sup>∗</sup> , w}<sub>∗</sub></formula>with<formula xmlns="http://xml-maiden.com">h<sup>∗</sup> = −¼ <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>vwdx</formula>So the non-Noether symmetry of Broer-Kaup system yields infinite sequenceof conservation laws of Broer-Kaup hierarchy and endows it with bi-Hamiltonian structure.</div><div class="paragraph">By suppressing dispersive terms in Broer-Kaup system one reduces it to more simpleintegarble model — dispersiveless long wave system <a href="#r73">[73]</a>,<a href="#r74">[74]</a><formula xml:id="e270" xmlns="http://xml-maiden.com">v<sub>t</sub> = v<sub>x</sub>w + vw<sub>x</sub><line/>w<sub>t</sub> = v<sub>x</sub> + ww<sub>x</sub></formula>in this case symmetry <a href="#e248">(248)</a> reduces to more simple non-Noether symmetry<formula xml:id="e271" xmlns="http://xml-maiden.com">E(v) = 4vw + 2x(vw)<sub>x</sub> + 3t(v<sup>2</sup> + vw<sup>2</sup>)<sub>x</sub><line/>E(w) = w<sup>2</sup> + 4v + 2x(ww<sub>x</sub> + v<sub>x</sub>) + t(6vw + w<sup>3</sup>)<sub>x</sub></formula>while the conservation laws of Broer-Kaup hierarchy reduce tosequence of conservation laws of dispersiveless long wave system<formula xml:id="e272" xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> wdx<line/>J<sup>(1)</sup> = L<sub>E</sub>J<sup>(0)</sup> = 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vdx<line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> = (L<sub>E</sub>)<sup>2</sup>J<sup>(0)</sup> =4 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vwdx<line/>J<sup>(3)</sup> = L<sub>E</sub>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>3</sup>J<sup>(0)</sup> =12 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (vw<sup>2</sup> + v<sup>2</sup>)dx<line/>J<sup>(4)</sup> = L<sub>E</sub>J<sup>(3)</sup> = (L<sub>E</sub>)<sup>4</sup>J<sup>(0)</sup> =48 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (3v<sup>2</sup>w + vw<sup>3</sup>)dx<line/>J<sup>(n)</sup> = L<sub>E</sub>J<sup>(n − 1)</sup> = (L<sub>E</sub>)<sup>n</sup>J<sup>(0)</sup></formula></div><div class="paragraph">At the same time bi-Hamitonian structure of Broer-Kaup hierarchy, after reductiongives rise to bi-Hamiltonian structure of dispersiveless long wave system<formula xml:id="e273" xmlns="http://xml-maiden.com">W = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> {<fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx<line/>Ŵ = [E , W] = − 2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> {v <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub><line/>+ w <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub> + <fraction><num>δ</num><den>δw</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx</formula></div><div class="paragraph">Among other reductions of dispersive water wave system one should probably mentionBurger's equation <a href="#r73">[73]</a>,<a href="#r74">[74]</a><formula xml:id="e274" xmlns="http://xml-maiden.com">w<sub>t</sub> = w<sub>xx</sub> + ww<sub>x</sub></formula>However Hamiltonian realization of this equation is unknown(for instance Poisson bivector field of dispersive water wave system<a href="#e252">(252)</a> vanishes during reduction).</div><h2>Benney system</h2><div class="paragraph">Now let us consider another integrable system of nonlinear partialdifferential equations — Benney system <a href="#r73">[73]</a>,<a href="#r74">[74]</a>. Time evolution of this dynamicalsystem is governed by equations of motion <formula xml:id="e275" xmlns="http://xml-maiden.com">u<sub>t</sub> = vv<sub>x</sub> + 2(uw)<sub>x</sub><line/>v<sub>t</sub> = 2u<sub>x</sub> + (vw)<sub>x</sub><line/>w<sub>t</sub> = 2v<sub>x</sub> + 2ww<sub>x</sub></formula>To determine symmetries of the system one has to look for solutions oflinear equation<formula xml:id="e276" xmlns="http://xml-maiden.com">E(u)<sub>t</sub> = (vE(v))<sub>x</sub> + 2(uE(w))<sub>x</sub> + 2(wE(u))<sub>x</sub><line/>E(v)<sub>t</sub> = 2E(u)<sub>x</sub> + (vE(w))<sub>x</sub> + (wE(v))<sub>x</sub><line/>E(w)<sub>t</sub> = 2E(v)<sub>x</sub> + 2(wE(w))<sub>x</sub></formula>obtained by substituting infinitesimal transformations<formula xmlns="http://xml-maiden.com">u  →  u + zE(u) + O(z<sup>2</sup>)<line/>v  →  v + zE(v) + O(z<sup>2</sup>)<line/>w  →  w + zE(w) + O(z<sup>2</sup>)</formula>into equations <a href="#e275">(275)</a> and grouping first order terms.In particular one can check that the vector field <math xmlns="http://xml-maiden.com">E</math> defined by<formula xml:id="e278" xmlns="http://xml-maiden.com">E(u) = 5uw + 2v<sup>2</sup> + x(2(uw)<sub>x</sub> + vv<sub>x</sub>) + 2t(4uv + v<sup>2</sup>w + 3uw<sup>2</sup>)<sub>x</sub><line/>E(v) = vw + 6u + x((vw)<sub>x</sub> + 2u<sub>x</sub>) + 2t(4uw + 3v<sup>2</sup> + vw<sup>2</sup>)<sub>x</sub><line/>E(w) = w<sup>2</sup> + 4v + 2x(ww<sub>x</sub> + v<sub>x</sub>) + 2t(w<sup>3</sup> + 4vw + 4u)<sub>x</sub></formula>satisfies equation <a href="#e276">(276)</a> and therefore generates symmetry of Benney system.The fact that this symmetry is local simplifies further calculations.</div><div class="paragraph">At the same time, it is known fact, that under zero boundary conditions<formula xmlns="http://xml-maiden.com">u(± ∞) = v(± ∞) = w(± ∞) = 0</formula>Benney equations can be rewritten in Hamiltonian form<formula xmlns="http://xml-maiden.com">u<sub>t</sub> = {h , u}<line/>v<sub>t</sub> = {h , v}<line/>w<sub>t</sub> = {h , w}</formula>with Hamiltonian<formula xml:id="e281" xmlns="http://xml-maiden.com">h = − ½ <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (2uw<sup>2</sup> + 4uv + v<sup>2</sup>w)dx</formula>and Poisson bracket defined by the following Poisson bivector field<formula xml:id="e282" xmlns="http://xml-maiden.com">W = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>{½ <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub>+ <fraction><num>δ</num><den>δu</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx</formula>Using symmetry <a href="#e278">(278)</a> that in fact is non-Noether one, we can reproducesecond Poisson bivector field involved in the bi-Hamiltonian structure of Benney hierarchy(by taking Lie derivative of <math xmlns="http://xml-maiden.com">W</math> along <math xmlns="http://xml-maiden.com">E</math>)<formula xml:id="e283" xmlns="http://xml-maiden.com">Ŵ = [E , W] = − 3 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>u <fraction><num>δ</num><den>δu</den></fraction> ∧ {<fraction><num>δ</num><den>δu</den></fraction>}<sub>x</sub>+ v <fraction><num>δ</num><den>δu</den></fraction> ∧ {<fraction><num>δ</num><den>δv</den></fraction>}<sub>x</sub><line/>+ w <fraction><num>δ</num><den>δu</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>+ 2 <fraction><num>δ</num><den>δv</den></fraction> ∧ {<fraction><num>δ</num><den>δw</den></fraction>}<sub>x</sub>} dx</formula>Poisson bracket defined by bivector field <math xmlns="http://xml-maiden.com">Ŵ</math> gives riseto the second Hamiltonian realization of Benney system<formula xmlns="http://xml-maiden.com">u<sub>t</sub> = {h<sup>∗</sup> , u}<sub>∗</sub><line/>v<sub>t</sub> = {h<sup>∗</sup> , v}<sub>∗</sub><line/>w<sub>t</sub> = {h<sup>∗</sup> , w}<sub>∗</sub></formula>with new Hamiltonian<formula xmlns="http://xml-maiden.com">h<sup>∗</sup> = <fraction><num>1</num><den>6</den></fraction> <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (v<sup>2</sup> + 2uw)dx</formula>Thus symmetry <a href="#e278">(278)</a> is closely related tobi-Hamiltonian realization of Benney hierarchy.</div><div class="paragraph">The same symmetry yields infinite sequence of conservation laws of Benney system.Namely one can construct sequence of integrals of motion by applying non-Noethersymmetry <a href="#e278">(278)</a> to<formula xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under>wdx</formula>(the fact that <math xmlns="http://xml-maiden.com">J<sup>(0)</sup></math> is conserved can be verified by integratingthird equation of Benney system). The sequence looks like<formula xml:id="e287" xmlns="http://xml-maiden.com">J<sup>(0)</sup> = <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> wdx<line/>J<sup>(1)</sup> = L<sub>E</sub>J<sup>(0)</sup> =2 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> vdx<line/>J<sup>(2)</sup> = L<sub>E</sub>J<sup>(1)</sup> = (L<sub>E</sub>)<sup>2</sup>J<sup>(0)</sup> =8 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> udx<line/>J<sup>(3)</sup> = L<sub>E</sub>J<sup>(2)</sup> = (L<sub>E</sub>)<sup>3</sup>J<sup>(0)</sup> =12 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (v<sup>2</sup> + 2uw)dx<line/>J<sup>(4)</sup> = L<sub>E</sub>J<sup>(3)</sup> = (L<sub>E</sub>)<sup>4</sup>J<sup>(0)</sup> =48 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (2uw<sup>2</sup> + 4uv + v<sup>2</sup>w)dx<line/>J<sup>(5)</sup> = L<sub>E</sub>J<sup>(4)</sup> = (L<sub>E</sub>)<sup>5</sup>J<sup>(0)</sup> =240 <under><over><sat>+ ∞</sat><ope>∫</ope></over><sat>− ∞</sat></under> (4u<sup>2</sup> + 8uvw + 2uw<sup>3</sup> + 2v<sup>3</sup> + v<sup>2</sup>w<sup>2</sup>)dx<line/>J<sup>(n)</sup> = L<sub>E</sub>J<sup>(n − 1)</sup> = (L<sub>E</sub>)<sup>n</sup>J<sup>(0)</sup></formula>So conservation laws and bi-Hamiltonian structure of Benney hierarchyare closely related to its symmetry, that can play important role in analysis ofBenney system and other models that can be obtained from it by reduction.</div><h2>Conclusions</h2><div class="paragraph">The fact that many important integrable models, such as Korteweg-de Vries equation, Broer-Kaup system, Benney system and Toda chain,possess non-Noether symmetries that can be effectively usedin analysis of these models, inclines us to think that non-Noether symmetries can playessential role in theory of integrable systems and properties of this class of symmetriesshould be investigated further.The present review indicates that in many cases non-Noether symmetries lead to maximal involutivefamilies of functionally independent conserved quantities and in this way ensure integrabilityof dynamical system. To determine involutivity of conservation laws in cases when it can not be checkedby direct computations (for instance  one can not check directly the involutivityin many generic n-dimensional models like Toda chainand infinite dimensional models like KdV hierarchy)we propose analog of Yang-Baxter equation, that being satisfied bygenerator of symmetry, ensures involutivity of familyof conserved quantities associated with this symmetry.</div><div class="paragraph">Another important feature of non-Noether symmetries is their relationship withseveral essential geometric concepts, emerging in theory of integrable systems, such asFrölicher-Nijenhuis operators, Lax pairs, bi-Hamiltonian structures andbicomplexes. On the one hand this relationship enlarges possible scope ofapplications of non-Noether symmetries in Hamiltonian dynamics and on the other hand itindicates that existence of invariant Frölicher-Nijenhuis operators,bi-Hamiltonian structures and bicomplexes in many cases can be considered as manifestationof hidden symmetries of dynamical system.</div><div class="acknowledgements"><strong class="cap">acknowledgements. </strong>Author is grateful to George Jorjadze, Zakaria Giunashvili andMichael Maziashvili for constructive discussions and help.This work was supported by INTAS (00-00561).</div><h2 class="references">References</h2><ol class="references"><li>	<span class="who">M. Antonowicz, A. Fordy</span>	<span class="what">Coupled KdV equations with multi-Hamiltonian structures</span>	<span class="where">Physica D, 28 345-357</span>	<span class="when">1987</span></li><li>	<span class="who">V. Arnold</span>	<span class="what">Dynamical systems III, 2nd ed.</span>	<span class="where">EMS, Springer-Verlag</span>	<span class="when">1993</span></li><li>	<span class="who">V. Arnold, S. Novikov</span>	<span class="what">Dynamical systems IV, 2nd ed.</span>	<span class="where">EMS, Springer-Verlag</span>	<span class="when">2001</span></li><li>	<span class="who">G. Bluman, S. Kumei</span>	<span class="what">Symmetries and differential equations</span>	<span class="where">Springer-Verlag, New York</span>	<span class="when">1989</span></li><li>	<span class="who">A. Bocharov et al.</span>	<span class="what">Symmetries and conservation laws for differential equations of mathematical physics</span>	<span class="where">AMS, TMM v. 182</span>	<span class="when">1999</span></li><li>	<span class="who">R. Brouzet</span>	<span class="what">Sur quelques propriétés géométriques des variétés bihamiltoniennes</span>	<span class="where">C. R. Acad. Sci. Paris 308, série I, 287-92</span>	<span class="when">1989</span></li><li>	<span class="who">R. Brouzet</span>	<span class="what">Systèmes biHamiltonienes et complète intégrabilité en dimension 4</span>	<span class="where">Comptes Rendus Acad. Sci. Paris 311 895-898</span>	<span class="when">1990</span></li><li>	<span class="who">J. Cariñena, L. Ibort</span>	<span class="what">Non-Noether constants of motion</span>	<span class="where">J. Phys. A:Math. Gen. 16, no. 1, 1-7</span>	<span class="when">1983</span></li><li>	<span class="who">J. Cariñena, G. Marmo, M. Rañada</span>	<span class="what">Non-symplectic symmetries and bi-Hamiltonian structures of rational harmonic oscillator</span>	<span class="where">J. of Phys. A: Math. Gen. 35 L679-686</span>	<span class="when">2002</span></li><li>	<span class="who">P. Casati, F. Magri, M. Pedroni</span>	<span class="what">BiHamiltonian manifolds and Sato's equations</span>	<span class="where">in O. Babelon et al., Integrable systems</span>	<span class="when">1993</span></li><li>	<span class="who">G. Chavchanidze</span>	<span class="what">Non-Noether symmetries in integrable models</span>	<span class="where">J. Phys. A: Math. Gen. 37  2253-2260, math-ph/0307018</span>	<span class="when">2004</span></li><li>	<span class="who">G. Chavchanidze</span>	<span class="what">Non-Noether symmetries and their influence on phase space geometry</span>	<span class="where">J. Geom. Phys. 48, 190-202</span>	<span class="when">2003</span></li><li>	<span class="who">G. Chavchanidze, </span>	<span class="what">Bi-Hamiltonian structure as a shadow of non-Noether symmetry</span>	<span class="where">Georgian Math. J. 10, 057-061</span>	<span class="when">2003</span></li><li>	<span class="who">G. Chavchanidze</span>	<span class="what">Non-Noether symmetries in singular dynamical systems</span>	<span class="where">Georgian Math. J. 8 (2001) 027-032</span>	<span class="when">2001</span></li><li>	<span class="who">A. Chowdhury</span>	<span class="what">Lie algebraic methods in integrable systems</span>	<span class="where"> Longman Group</span>	<span class="when">2000</span></li><li>	<span class="who">M. Crampin</span>	<span class="what">A note on non-Noether constants of motion</span>	<span class="where">Phys. Lett. A, 95, 209-212</span>	<span class="when">1983</span></li><li>	<span class="who">M. Crampin, W. Sarlet, G. Thompson</span>	<span class="what">Bi-differential calculi and bi-Hamiltonian systems</span>	<span class="where">J. of Phys. A: Math. Gen. 33, L177-180</span>	<span class="when">2000</span></li><li>	<span class="who">M. Crampin, W. Sarlet, G. Thompson</span>	<span class="what">Bi-differential calculi, bi-Hamiltonian systems and conformal Killing tensors</span>	<span class="where">J. of Phys. A: Math. Gen. 33, 8755-8770</span>	<span class="when">2000</span></li><li>	<span class="who">P. Damianou</span>	<span class="what">Multiple Hamiltonian structures for Toda-type systems</span>	<span class="where">J. Math. Phys. 35 (10) 5511-5541</span>	<span class="when">1994</span></li><li>	<span class="who">P. Damianou</span>	<span class="what">Symmetries of Toda equations</span>	<span class="where">J. of Phys. A: Math. Gen. 26, 3791-3796</span>	<span class="when">1993</span></li><li>	<span class="who">P. Damianou, C. Sophocleous</span>	<span class="what">Master and Noether symmetries for the Toda lattice</span>	<span class="where">Proceedings of the 16th International Symposium on Nonlinear Acoustics, vol 1, 618-622 </span>	<span class="when">2003</span></li><li>	<span class="who">A. Das</span>	<span class="what">Integrable models</span>	<span class="where">World Scientific Lecture Notes in Physics, Vol. 30</span>	<span class="when">1989</span></li><li>	<span class="who">L. Dickey</span>	<span class="what">Soliton equations and Hamiltonian systems</span>	<span class="where">World Scientific</span>	<span class="when">2003</span></li><li>	<span class="who">A. Dimakis, F. Müller-Hoissen</span>	<span class="what">Bicomplexes and integrable models</span>	<span class="where">nlin.SI/0006029</span>	<span class="when">2000 </span></li><li>	<span class="who">A. Dimakis, F. Müller-Hoissen</span>	<span class="what">Bi-differential calculi and integrable models</span>	<span class="where">J. of Phys. A: Math. Gen. 33 210-23</span>	<span class="when">2000</span></li><li>	<span class="who">V. Drinfeld, V. Sokolov</span>	<span class="what">Lie algebras and equations of Korteweg-de Vries type</span>	<span class="where">J. Sov. Math. 30, 1975-2036</span>	<span class="when">1985</span></li><li>	<span class="who">G. Falqui, F. Magri, M. Pedroni</span>	<span class="what">The method of Poisson pairs in the theory of nonlinear PDEs</span>	<span class="where">nlin.SI/0002009</span>	<span class="when">2000 </span></li><li>	<span class="who">G. Falqui, F. Magri, M. Pedroni</span>	<span class="what">Bihamiltonian geometry and separation of variables for Toda lattices</span>	<span class="where">J. Nonlinear Math. Phys. 8, suppl., 118-127</span>	<span class="when">2001</span></li><li>	<span class="who">G. Falqui, M. Pedroni</span>	<span class="what">Separation of variables for bi-Hamiltonian systems</span>	<span class="where">Math. Phys. Anal. Geom. 6, no. 2, 139-179</span>	<span class="when">2003</span></li><li>	<span class="who">R. Fernandes</span>	<span class="what">On the master symmetries and bi-Hamiltonian structure of the Toda lattice</span>	<span class="where">J. of Phys. A: Math. Gen. 26, 3793-3803</span>	<span class="when">1993</span></li><li>	<span class="who">R. Fernandes</span>	<span class="what">Completely integrable biHamiltonian systems</span>	<span class="where">University of Minnesota</span>	<span class="when">1991</span></li><li>	<span class="who">A. Fokas</span>	<span class="what">Generalized symmetries and constants of motion of evolution equations</span>	<span class="where">Lett. Math. Phys. 3, 1318-1325</span>	<span class="when">1979</span></li><li>	<span class="who">B. Fuchssteiner, A. Fokas</span>	<span class="what">Symplectic structures, their Bäcklund transformations and heredity symmetries</span>	<span class="where">Physica 4D, 47-66</span>	<span class="when">1981</span></li><li>	<span class="who">S. Ghosh, B. Talukdar, J. Shamanna</span>	<span class="what">On the bi-Hamiltonian structure of the KdV hierarchy</span>	<span class="where">Czechoslovak J. Phys. 53, no. 5, 425-432</span>	<span class="when">2003</span></li><li>	<span class="who">G. Giachetta, L. Mangiarotti, G. Sardanashvily</span>	<span class="what">Bi-Hamiltonian partially integrable systems</span>	<span class="where">J. Math. Phys. 44, no. 5, 1984-1997</span>	<span class="when">2003</span></li><li>	<span class="who">F. González-Gascón</span>	<span class="what">Geometric foundations of a new conservation law discovered by Hojman</span>	<span class="where">J. of Phys. A: Math. Gen. 27 L59-60</span>	<span class="when">1994 </span></li><li>	<span class="who">P. Guha</span>	<span class="what">A Note on Bidifferential Calculi and Bihamiltonian systems</span>	<span class="where">IHÉS preprint M/64</span>	<span class="when">2000 </span></li><li>	<span class="who">R. Hermann</span>	<span class="what">Topics in the geometric theory of integrable mechanical systems</span>	<span class="where">Math SCI Press</span>	<span class="when">1984</span></li><li>	<span class="who">S. Hojman</span>	<span class="what">A new conservation law constructed without using either Lagrangians or Hamiltonians</span>	<span class="where">J. of Phys. A: Math. Gen. 25 L291-295</span>	<span class="when">1992 </span></li><li>	<span class="who">S. Hojman, H. Harleston</span>	<span class="what">Equivalent Lagrangians: multi-dimensional case</span>	<span class="where">J. Math. Phys., 22, 1414-1419</span>	<span class="when">1981</span></li><li>	<span class="who">J. Hoppe</span>	<span class="what">Lectures on integrable systems</span>	<span class="where">Springer-Verlag</span>	<span class="when">1992</span></li><li>	<span class="who">A. Ibort, F. Magri, G. Marmo</span>	<span class="what">Bihamiltonian structures and Stäckel separability</span>	<span class="where">J. Geom. Phys. 32, 210-228</span>	<span class="when">2000</span></li><li>	<span class="who">N. Ibragimov</span>	<span class="what">Transformation groups applied to mathematical physics</span>	<span class="where">Reidel, Boston</span>	<span class="when">1985</span></li><li>	<span class="who">Y. Kosmann-Schwarzbach</span>	<span class="what">Generalized symmetries of nonlinear partial differential equations</span>	<span class="where">Lett. Math. Phys. 3, 395-404</span>	<span class="when">1979</span></li><li>	<span class="who">Y. Kosmann-Schwarzbach, F. Magri</span>	<span class="what">Poisson-Nijenhuis structures</span>	<span class="where">Ann. Inst. Poincaré, 53, 35-81</span>	<span class="when">1990</span></li><li>	<span class="who">M. Leo, R. Leo, G. Soliani, L. Solombrino, G. Mancarella</span>	<span class="what">Symmetry properties and bi-Hamiltonian structure of the Toda lattice</span>	<span class="where">Lett. Math. Phys. 8, 267-272</span>	<span class="when">1984</span></li><li>	<span class="who">P. Libermann, C-M. Marle</span>	<span class="what">Symplectic geometry and analytical mechanics</span>	<span class="where">Reidel, Dordrecht-Boston</span>	<span class="when">1987</span></li><li>	<span class="who">C. López, E. Martinez, M. Rañada</span>	<span class="what">Dynamical symmetries, non-Cartan symmetries and superintegrability of the n-dimensional harmonic oscillator</span>	<span class="where">J. of Phys. A: Math. Gen. 32, 1241-1249</span>	<span class="when">1999</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Conserved quantities from non-Noether symmetries without alternative Lagrangians</span>	<span class="where">Internat. J. Non-Linear Mech. 34, no. 2, 387-390</span>	<span class="when">1999</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Conserved quantities and velocity-dependent symmetries in Lagrangian dynamics</span>	<span class="where">Internat. J. Non-Linear Mech. 33, no. 2, 393-396</span>	<span class="when">1998</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">New derivation of a conserved quantity for Lagrangian systems</span>	<span class="where">J. of Phys. A: Math. Gen. 15, L721-722</span>	<span class="when">1998</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Remarks on a recent theorem about conserved quantities</span>	<span class="where">J. of Phys. A: Math. Gen. 28, L637-638</span>	<span class="when">1995</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Noncannonical symmetries, bi-Hamiltonian systems and complete integrablity</span>	<span class="where">J. Phys. Soc. Japan 54, 64-68</span>	<span class="when">1985</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">New classes of conserved quantities associated with non-Noether symmetries</span>	<span class="where">J. Phys. A 15, no. 3, L87-L91</span>	<span class="when">1982</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Isospectral representations of dynamical systems</span>	<span class="where">Lett. Nuovo Cim. 35, 393-397</span>	<span class="when">1982</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Noncannonical symmetries and isospectral representations of Hamiltonian systems</span>	<span class="where">Phys. Lett. 87A, 274-276</span>	<span class="when">1982</span></li><li>	<span class="who">M. Lutzky</span>	<span class="what">Origin of non-Noether invariants</span>	<span class="where">Phys. Lett. A 75, no. 1-2, 8-10</span>	<span class="when">1980</span></li><li>	<span class="who">F. Magri</span>	<span class="what">A simple model of the integrable Hamiltonian equation</span>	<span class="where">J. Math. Phys. 19 no.5, 1156-1162</span>	<span class="when">1978</span></li><li>	<span class="who">C-M. Marle, J. Nunes da Costa</span>	<span class="what">Master symmetries and bi-Hamiltonian structures for the relativistic Toda lattice</span>	<span class="where">J. of Phys. A: Math. Gen. 30, 7551-7556</span>	<span class="when">1997</span></li><li>	<span class="who">J. Marsden, T. Ratiu</span>	<span class="what">Introduction to mechanics and symmetry</span>	<span class="where">Springer-Verlag</span>	<span class="when">1999</span></li><li>	<span class="who">C. Morosi, G. Tondo</span>	<span class="what">Quasi-bihamiltonian systems and separability</span>	<span class="where">J. of Phys. A: Math. Gen. 30, 2799-2806</span>	<span class="when">1997</span></li><li>	<span class="who">W. Oevel</span>	<span class="what">Topics in soliton theory and exactly solvable non-linear equations</span>	<span class="where">World Scientific</span>	<span class="when">1987</span></li><li>	<span class="who">W. Oevel, B. Fuchssteiner, H. Zhang, O. Ragnisco</span>	<span class="what">Mastersymmetries, angle variables, and recursion operator of the relativistic Toda lattice</span>	<span class="where">J. Math. Phys. 30 2664-2670 </span>	<span class="when">1989</span></li><li>	<span class="who">P. Olver</span>	<span class="what">Cannonical forms of biHamiltonian systems</span>	<span class="where">in O. Babelon et al., Integrable systems</span>	<span class="when">1993</span></li><li>	<span class="who">P. Olver</span>	<span class="what">Cannonical forms and integrability of biHamiltonian systems</span>	<span class="where">Phys. Lett. 148 A, 177-187</span>	<span class="when">1990</span></li><li>	<span class="who">P. Olver</span>	<span class="what">Applications of Lie groups to differential equations</span>	<span class="where">GTM 107, Springer Verlag, New York</span>	<span class="when">1986</span></li><li>	<span class="who">M. Rañada</span>	<span class="what">Dynamical symmetries, bi-Hamiltonian structures and superintegrable n = 2 systems</span>	<span class="where">J. Math. Phys. 41 (4) 2121-2134</span>	<span class="when">2000</span></li><li>	<span class="who">A. Razumov, M. Saveliev</span>	<span class="what">Lie algebras, geometry and Toda-type systems</span>	<span class="where">Cambridge University Press</span>	<span class="when">1997</span></li><li>	<span class="who">C. Rogers, W. Schief</span>	<span class="what">Bäcklund and Darboux transformations</span>	<span class="where">Cambridge University Press</span>	<span class="when">2002</span></li><li>	<span class="who">A. Sergyeyev</span>	<span class="what">A simple way of making a Hamiltonian system into a bi-Hamiltonian one</span>	<span class="where">Acta Appl. Math. 83, 183-197</span>	<span class="when">2004</span></li><li>	<span class="who">Y. Suris</span>	<span class="what">On the bi-Hamiltonian structure of Toda and relativistic Toda lattices</span>	<span class="where">Phys. Lett. A 180 419 </span>	<span class="when">1993</span></li><li>	<span class="who">F. Turiel</span>	<span class="what">Classification locale d'un couple de formes symplectiques Poisson-compatibles</span>	<span class="where">Comptes Rendus Acad. Sci. Paris 308, 575-578</span>	<span class="when">1989</span></li><li>	<span class="who">J.P. Wang</span>	<span class="what">A list of 1 + 1 dimensional integrable equations and their properties</span>	<span class="where">J. Nonlinear Math. Phys. 9, suppl. 1, 213–233</span>	<span class="when">2002</span></li><li>	<span class="who">J.P. Wang</span>	<span class="what">Symmetries and conservation laws of evolution equations</span>	<span class="where">Ph.D. Thesis, Vrije Universiteit van Amsterdam</span>	<span class="when">1998</span></li><li>	<span class="who">A. Weinstein</span>	<span class="what">The local structure of Poisson manifolds</span>	<span class="where">J. Diff. Geom. 18, 523-557</span>	<span class="when">1983</span></li></ol></body></html>