Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

On the field analogue of elliptic spin Calogero-Moser model:

Lax pair and equations of motion

Andrei Zotov

Steklov Mathematical Institute of Russian Academy of Sciences,
Gubkina str. 8, 119991, Moscow, Russia

Institute for Theoretical and Mathematical Physics,
Lomonosov Moscow State University, Moscow, 119991, Russia

e-mail: zotov@mi-ras.ru

Abstract

The Lax pair for the field analogue of the classical spin elliptic Calogero-Moser is proposed. Namely, using the previously known Lax matrix we suggest an ansatz for the accompany matrix. The presented construction is valid when the matrix of spin variables ๐’ฎโˆˆMatโข(N,โ„‚)๐’ฎMat๐‘โ„‚{\mathcal{S}}\in{\rm Mat}(N,\mathbb{C})caligraphic_S โˆˆ roman_Mat ( italic_N , blackboard_C ) satisfies the condition ๐’ฎ2=c0โข๐’ฎsuperscript๐’ฎ2subscript๐‘0๐’ฎ{\mathcal{S}}^{2}=c_{0}{\mathcal{S}}caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S with some constant c0โˆˆโ„‚subscript๐‘0โ„‚c_{0}\in\mathbb{C}italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT โˆˆ blackboard_C. It is proved that the Lax pair satisfies the Zakharov-Shabat equation with unwanted term, thus providing equations of motion on the unreduced phase space. The unwanted term vanishes after additional reduction. In the special case rankโข(๐’ฎ)=1rank๐’ฎ1{\rm rank}(\mathcal{S})=1roman_rank ( caligraphic_S ) = 1 we show that the reduction provides the Lax pair of the spinless field Calogero-Moser model obtained earlier by Akhmetshin, Krichever and Volvovski.


1 Introduction

The spinless elliptic Calogero-Moser model [5] (of ANโˆ’1subscript๐ด๐‘1A_{N-1}italic_A start_POSTSUBSCRIPT italic_N - 1 end_POSTSUBSCRIPT type) is an example of N๐‘Nitalic_N-body integrable system of classical finite-dimensional mechanics. Dynamics of particles is defined by the equations of motion

qห™i=b~โขpi,pห™i=b~โขc2โขโˆ‘k:kโ‰ iNโ„˜โ€ฒโข(qiโˆ’qk),i=1,โ€ฆ,N,formulae-sequencesubscriptห™๐‘ž๐‘–~๐‘subscript๐‘๐‘–formulae-sequencesubscriptห™๐‘๐‘–~๐‘superscript๐‘2superscriptsubscript:๐‘˜๐‘˜๐‘–๐‘superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–subscript๐‘ž๐‘˜๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{{\dot{q}}_{i}={\tilde{b}}p_{i}\,,\qquad{\dot{p% }}_{i}={\tilde{b}}{c}^{2}\sum\limits_{k:k\neq i}^{N}\wp^{\prime}(q_{i}-q_{k})% \,,\quad i=1,...,N\,,}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , overห™ start_ARG italic_p end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_k : italic_k โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) , italic_i = 1 , โ€ฆ , italic_N , end_CELL end_ROW end_ARRAY (1.1)

where pisubscript๐‘๐‘–p_{i}italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, qisubscript๐‘ž๐‘–q_{i}italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, i=1,โ€ฆ,N๐‘–1โ€ฆ๐‘i=1,...,Nitalic_i = 1 , โ€ฆ , italic_N are canonically conjugated โ„‚โ„‚\mathbb{C}blackboard_C-valued momenta and positions of particles111The Poisson brackets are {pi,qj}=ฮดiโขjsubscript๐‘๐‘–subscript๐‘ž๐‘—subscript๐›ฟ๐‘–๐‘—\{p_{i},q_{j}\}=\delta_{ij}{ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT } = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT, {pi,pj}={qi,qj}=0subscript๐‘๐‘–subscript๐‘๐‘—subscript๐‘ž๐‘–subscript๐‘ž๐‘—0\{p_{i},p_{j}\}=\{q_{i},q_{j}\}=0{ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT } = { italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT } = 0, cโˆˆโ„‚๐‘โ„‚c\in\mathbb{C}italic_c โˆˆ blackboard_C is a coupling constant, โ„˜โข(x)Weierstrass-p๐‘ฅ\wp(x)โ„˜ ( italic_x ) is the Weierstrass elliptic โ„˜Weierstrass-p\wpโ„˜-function222See Appendix for elliptic function definitions and properties and b~โ‰ 0~๐‘0{\tilde{b}}\neq 0over~ start_ARG italic_b end_ARG โ‰  0 is a free โ„‚โ„‚\mathbb{C}blackboard_C-valued normalization constant. It was shown by I. Krichever that equations (1.1) are represented in the Lax form

Lห™โข(z)+[Lโข(z),Mโข(z)]=0,Lโข(z),Mโข(z)โˆˆMatโข(N,โ„‚),formulae-sequenceห™๐ฟ๐‘ง๐ฟ๐‘ง๐‘€๐‘ง0๐ฟ๐‘ง๐‘€๐‘งMat๐‘โ„‚\begin{array}[]{c}\displaystyle{\dot{L}(z)+[L(z),M(z)]=0\,,\qquad L(z),M(z)\in% {\rm Mat}(N,\mathbb{C})\,,}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_L end_ARG ( italic_z ) + [ italic_L ( italic_z ) , italic_M ( italic_z ) ] = 0 , italic_L ( italic_z ) , italic_M ( italic_z ) โˆˆ roman_Mat ( italic_N , blackboard_C ) , end_CELL end_ROW end_ARRAY (1.2)

where zโˆˆโ„‚๐‘งโ„‚z\in\mathbb{C}italic_z โˆˆ blackboard_C is the spectral parameter. From (1.2) it follows that Hkโข(z)=trโข(Lkโข(z))subscript๐ป๐‘˜๐‘งtrsuperscript๐ฟ๐‘˜๐‘งH_{k}(z)={\rm tr}(L^{k}(z))italic_H start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ( italic_z ) = roman_tr ( italic_L start_POSTSUPERSCRIPT italic_k end_POSTSUPERSCRIPT ( italic_z ) ), kโˆˆโ„ค+๐‘˜subscriptโ„คk\in\mathbb{Z}_{+}italic_k โˆˆ blackboard_Z start_POSTSUBSCRIPT + end_POSTSUBSCRIPT are generating functions of integrals of motion. Existence of the classical r๐‘Ÿritalic_r-matrix structure guarantees the involution property {Hkโข(z),Hmโข(w)}=0subscript๐ป๐‘˜๐‘งsubscript๐ป๐‘š๐‘ค0\{H_{k}(z),H_{m}(w)\}=0{ italic_H start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ( italic_z ) , italic_H start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT ( italic_w ) } = 0, k,mโˆˆโ„ค+๐‘˜๐‘šsubscriptโ„คk,m\in\mathbb{Z}_{+}italic_k , italic_m โˆˆ blackboard_Z start_POSTSUBSCRIPT + end_POSTSUBSCRIPT. In this way the classical Liouville integrability of the model was established.

The Calogero-Moser model (1.1) has the spin generalization [7, 4]. In the classical mechanics this means existence of the โ€spinโ€ component of the phase space, which is a factor space ๐’ช//โ„‹{\mathcal{O}}//{\mathcal{H}}caligraphic_O / / caligraphic_H, where ๐’ช๐’ช{\mathcal{O}}caligraphic_O is a coadjoint orbit (of GLNsubscriptGL๐‘{\rm GL}_{N}roman_GL start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT Lie group) and โ„‹โŠ‚GLNโ„‹subscriptGLN{\mathcal{H}}\subset{\rm GL_{N}}caligraphic_H โŠ‚ roman_GL start_POSTSUBSCRIPT roman_N end_POSTSUBSCRIPT is the Cartan subgroup in GLNsubscriptGL๐‘{\rm GL}_{N}roman_GL start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT. Dynamical variables are elements of a matrix SโˆˆMatโข(N,โ„‚)๐‘†Mat๐‘โ„‚S\in{\rm Mat}(N,\mathbb{C})italic_S โˆˆ roman_Mat ( italic_N , blackboard_C ). The model is described on the unreduced phase space, that is when the spin component of the phase space is ๐’ช๐’ช{\mathcal{O}}caligraphic_O. This space is equipped with the linear Poisson-Lie structure {Siโขj,Skโขl}=โˆ’Siโขlโขฮดkโขj+Skโขjโขฮดiโขlsubscript๐‘†๐‘–๐‘—subscript๐‘†๐‘˜๐‘™subscript๐‘†๐‘–๐‘™subscript๐›ฟ๐‘˜๐‘—subscript๐‘†๐‘˜๐‘—subscript๐›ฟ๐‘–๐‘™\{S_{ij},S_{kl}\}=-S_{il}\delta_{kj}+S_{kj}\delta_{il}{ italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT , italic_S start_POSTSUBSCRIPT italic_k italic_l end_POSTSUBSCRIPT } = - italic_S start_POSTSUBSCRIPT italic_i italic_l end_POSTSUBSCRIPT italic_ฮด start_POSTSUBSCRIPT italic_k italic_j end_POSTSUBSCRIPT + italic_S start_POSTSUBSCRIPT italic_k italic_j end_POSTSUBSCRIPT italic_ฮด start_POSTSUBSCRIPT italic_i italic_l end_POSTSUBSCRIPT, for i,j,k,l=1,โ€ฆ,Nformulae-sequence๐‘–๐‘—๐‘˜๐‘™1โ€ฆ๐‘i,j,k,l=1,...,Nitalic_i , italic_j , italic_k , italic_l = 1 , โ€ฆ , italic_N, and the Casimir functions are fixed333Fixation of the Casimir functions means fixation of eigenvalues of the matrix S๐‘†Sitalic_S.. The equations of motion are as follows:

qห™i=b~โขpi,pห™i=b~โขโˆ‘j:jโ‰ iNSiโขjโขSjโขiโขโ„˜โ€ฒโข(qiโˆ’qj)formulae-sequencesubscriptห™๐‘ž๐‘–~๐‘subscript๐‘๐‘–subscriptห™๐‘๐‘–~๐‘superscriptsubscript:๐‘—๐‘—๐‘–๐‘subscript๐‘†๐‘–๐‘—subscript๐‘†๐‘—๐‘–superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{{\dot{q}}_{i}={\tilde{b}}p_{i}\,,\quad{\dot{p}% }_{i}={\tilde{b}}\sum\limits_{j:j\neq i}^{N}S_{ij}S_{ji}\wp^{\prime}(q_{i}-q_{% j})}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , overห™ start_ARG italic_p end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG โˆ‘ start_POSTSUBSCRIPT italic_j : italic_j โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_S start_POSTSUBSCRIPT italic_j italic_i end_POSTSUBSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (1.3)

and

Sห™iโขi=0,Sห™iโขj=b~โขโˆ‘kโ‰ i,jNSiโขkโขSkโขjโข(โ„˜โข(qiโˆ’qk)โˆ’โ„˜โข(qjโˆ’qk)),iโ‰ j.formulae-sequencesubscriptห™๐‘†๐‘–๐‘–0formulae-sequencesubscriptห™๐‘†๐‘–๐‘—~๐‘superscriptsubscript๐‘˜๐‘–๐‘—๐‘subscript๐‘†๐‘–๐‘˜subscript๐‘†๐‘˜๐‘—Weierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘˜Weierstrass-psubscript๐‘ž๐‘—subscript๐‘ž๐‘˜๐‘–๐‘—\begin{array}[]{c}\displaystyle{{\dot{S}}_{ii}=0\,,\quad{\dot{S}}_{ij}={\tilde% {b}}\sum\limits_{k\neq i,j}^{N}S_{ik}S_{kj}(\wp(q_{i}-q_{k})-\wp(q_{j}-q_{k}))% \,,\ i\neq j\,.}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = 0 , overห™ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG โˆ‘ start_POSTSUBSCRIPT italic_k โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_S start_POSTSUBSCRIPT italic_i italic_k end_POSTSUBSCRIPT italic_S start_POSTSUBSCRIPT italic_k italic_j end_POSTSUBSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) ) , italic_i โ‰  italic_j . end_CELL end_ROW end_ARRAY (1.4)

At the same time the Lax equation on the unreduced phase space has a certain nontrivial right hand side [4]:

Lห™โข(z)+[Lโข(z),Mโข(z)]=unwantedโขterm.ห™๐ฟ๐‘ง๐ฟ๐‘ง๐‘€๐‘งunwantedterm\begin{array}[]{c}\displaystyle{\dot{L}(z)+[L(z),M(z)]={\rm unwanted\ term}\,.% }\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_L end_ARG ( italic_z ) + [ italic_L ( italic_z ) , italic_M ( italic_z ) ] = roman_unwanted roman_term . end_CELL end_ROW end_ARRAY (1.5)

The unwanted term vanishes on the reduced phase space. For this purpose one should perform the Poisson reduction ๐’ชโ†’๐’ช//โ„‹{\mathcal{O}}\rightarrow{\mathcal{O}}//{\mathcal{H}}caligraphic_O โ†’ caligraphic_O / / caligraphic_H. It includes (the first class) constraints ฮผi=constsubscript๐œ‡๐‘–const\mu_{i}={\rm const}italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = roman_const, i=1,โ€ฆ,N๐‘–1โ€ฆ๐‘i=1,...,Nitalic_i = 1 , โ€ฆ , italic_N

ฮผi=Siโขi,i=1,โ€ฆ,Nformulae-sequencesubscript๐œ‡๐‘–subscript๐‘†๐‘–๐‘–๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{\mu_{i}=S_{ii}\,,\quad i=1,...,N}\end{array}start_ARRAY start_ROW start_CELL italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT , italic_i = 1 , โ€ฆ , italic_N end_CELL end_ROW end_ARRAY (1.6)

together with some (gauge) fixation of the coadjoint action of โ„‹โ„‹{\mathcal{H}}caligraphic_H on ๐’ช๐’ช{\mathcal{O}}caligraphic_O444 Conditions (1.6) arise as moment map equations generated by the coadjoint action of โ„‹โ„‹{\mathcal{H}}caligraphic_H on ๐’ช๐’ช{\mathcal{O}}caligraphic_O.. The reduction changes the Poisson brackets and equations of motion. However it is usually not performed explicitly since there is no any preferred way to fix the action of โ„‹โ„‹{\mathcal{H}}caligraphic_H. Notice also that the dimension of an orbit ๐’ช๐’ช{\mathcal{O}}caligraphic_O depends on the level of fixation of Specโข(S)Spec๐‘†{\rm Spec}(S)roman_Spec ( italic_S ). For example, in the minimal orbit case Specโข(S)=(c0,0,โ€ฆ,0)Spec๐‘†subscript๐‘00โ€ฆ0{\rm Spec}(S)=(c_{0},0,...,0)roman_Spec ( italic_S ) = ( italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT , 0 , โ€ฆ , 0 ), that is rankโข(S)=1rank๐‘†1{\rm rank}(S)=1roman_rank ( italic_S ) = 1 and the reduction kills all spin degrees of freedom reproducing the spinless model (1.1).

In the field case we deal with the field variables. The latter means that the momenta and coordinates of particles turn into the set of fields p1โข(x),โ€ฆ,pNโข(x)subscript๐‘1๐‘ฅโ€ฆsubscript๐‘๐‘๐‘ฅp_{1}(x),...,p_{N}(x)italic_p start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_x ) , โ€ฆ , italic_p start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ( italic_x ) and q1โข(x),โ€ฆ,qNโข(x)subscript๐‘ž1๐‘ฅโ€ฆsubscript๐‘ž๐‘๐‘ฅq_{1}(x),...,q_{N}(x)italic_q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_x ) , โ€ฆ , italic_q start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ( italic_x ), where xโˆˆโ„๐‘ฅโ„x\in\mathbb{R}italic_x โˆˆ blackboard_R is a coordinate on a real line or a circle. For definiteness we assume all the field variables be periodic functions on a circle ฯ†โข(x)=ฯ†โข(x+2โขฯ€)๐œ‘๐‘ฅ๐œ‘๐‘ฅ2๐œ‹\varphi(x)=\varphi(x+2\pi)italic_ฯ† ( italic_x ) = italic_ฯ† ( italic_x + 2 italic_ฯ€ ). The canonical Poisson structure takes the form {piโข(x),qjโข(y)}=ฮดโข(xโˆ’y)โขฮดiโขjsubscript๐‘๐‘–๐‘ฅsubscript๐‘ž๐‘—๐‘ฆ๐›ฟ๐‘ฅ๐‘ฆsubscript๐›ฟ๐‘–๐‘—\{p_{i}(x),q_{j}(y)\}=\delta(x-y)\delta_{ij}{ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_y ) } = italic_ฮด ( italic_x - italic_y ) italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT, {piโข(x),pjโข(y)}={qiโข(x),qjโข(y)}=0subscript๐‘๐‘–๐‘ฅsubscript๐‘๐‘—๐‘ฆsubscript๐‘ž๐‘–๐‘ฅsubscript๐‘ž๐‘—๐‘ฆ0\{p_{i}(x),p_{j}(y)\}=\{q_{i}(x),q_{j}(y)\}=0{ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) , italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_y ) } = { italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_y ) } = 0, x,yโˆˆโ„๐‘ฅ๐‘ฆโ„x,y\in\mathbb{R}italic_x , italic_y โˆˆ blackboard_R. The field analogue of the Calogero-Moser model was introduced in [9] and [10] for gl2subscriptgl2{\rm gl}_{2}roman_gl start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT case and then was extended to glNsubscriptgl๐‘{\rm gl}_{N}roman_gl start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT case in [1]. Equations of motion are given by the following set of partial differential equations for qiโข(t,x)subscript๐‘ž๐‘–๐‘ก๐‘ฅq_{i}(t,x)italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_t , italic_x ), piโข(t,x)subscript๐‘๐‘–๐‘ก๐‘ฅp_{i}(t,x)italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_t , italic_x ), i=1,โ€ฆ,N๐‘–1โ€ฆ๐‘i=1,...,Nitalic_i = 1 , โ€ฆ , italic_N:

qห™i=โˆ’2โข(kโขqi,x+c)โข(piโˆ’1Nโขcโขโˆ‘j=1Npjโข(kโขqj,x+c)),subscriptห™๐‘ž๐‘–2๐‘˜subscript๐‘ž๐‘–๐‘ฅ๐‘subscript๐‘๐‘–1๐‘๐‘superscriptsubscript๐‘—1๐‘subscript๐‘๐‘—๐‘˜subscript๐‘ž๐‘—๐‘ฅ๐‘\begin{array}[]{c}\displaystyle{{\dot{q}}_{i}=-2(kq_{i,x}+c)\Big{(}p_{i}-\frac% {1}{Nc}\sum\limits_{j=1}^{N}p_{j}(kq_{j,x}+c)\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = - 2 ( italic_k italic_q start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT + italic_c ) ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - divide start_ARG 1 end_ARG start_ARG italic_N italic_c end_ARG โˆ‘ start_POSTSUBSCRIPT italic_j = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_k italic_q start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT + italic_c ) ) , end_CELL end_ROW end_ARRAY (1.7)
pห™i=โˆ’kโขโˆ‚x(pi2โˆ’2โขpiโข1Nโขcโขโˆ‘j=1Npjโข(kโขqj,x+c)+12โขk3โขqi,xโขxโขxkโขqi,x+cโˆ’14โขk4โขqi,xโขx2(kโขqi,x+c)2)โˆ’โˆ’2โขโˆ‘j:jโ‰ iN((kโขqj,x+c)3โขโ„˜โ€ฒโข(qiโˆ’qj)โˆ’3โขk2โข(kโขqj,x+c)โขqj,xโขxโขโ„˜โข(qiโˆ’qj)โˆ’k3โขqj,xโขxโขxโขฮถโข(qiโˆ’qj)),subscriptห™๐‘๐‘–limit-from๐‘˜subscript๐‘ฅsuperscriptsubscript๐‘๐‘–22subscript๐‘๐‘–1๐‘๐‘superscriptsubscript๐‘—1๐‘subscript๐‘๐‘—๐‘˜subscript๐‘ž๐‘—๐‘ฅ๐‘12superscript๐‘˜3subscript๐‘ž๐‘–๐‘ฅ๐‘ฅ๐‘ฅ๐‘˜subscript๐‘ž๐‘–๐‘ฅ๐‘14superscript๐‘˜4subscriptsuperscript๐‘ž2๐‘–๐‘ฅ๐‘ฅsuperscript๐‘˜subscript๐‘ž๐‘–๐‘ฅ๐‘22superscriptsubscript:๐‘—๐‘—๐‘–๐‘superscript๐‘˜subscript๐‘ž๐‘—๐‘ฅ๐‘3superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–subscript๐‘ž๐‘—3superscript๐‘˜2๐‘˜subscript๐‘ž๐‘—๐‘ฅ๐‘subscript๐‘ž๐‘—๐‘ฅ๐‘ฅWeierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘—superscript๐‘˜3subscript๐‘ž๐‘—๐‘ฅ๐‘ฅ๐‘ฅ๐œsubscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{{\dot{p}}_{i}=-k\partial_{x}\bigg{(}p_{i}^{2}-% 2p_{i}\frac{1}{Nc}\sum\limits_{j=1}^{N}p_{j}(kq_{j,x}+c)+\frac{1}{2}\frac{k^{3% }q_{i,xxx}}{kq_{i,x}+c}-\frac{1}{4}\frac{k^{4}q^{2}_{i,xx}}{(kq_{i,x}+c)^{2}}% \bigg{)}-}\\ \displaystyle{-2\sum\limits_{j:j\neq i}^{N}\bigg{(}(kq_{j,x}+c)^{3}\wp^{\prime% }(q_{i}-q_{j})-3k^{2}(kq_{j,x}+c)q_{j,xx}\wp(q_{i}-q_{j})-k^{3}q_{j,xxx}\zeta(% q_{i}-q_{j})\bigg{)}\,,}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_p end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT - 2 italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT divide start_ARG 1 end_ARG start_ARG italic_N italic_c end_ARG โˆ‘ start_POSTSUBSCRIPT italic_j = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_k italic_q start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT + italic_c ) + divide start_ARG 1 end_ARG start_ARG 2 end_ARG divide start_ARG italic_k start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT italic_q start_POSTSUBSCRIPT italic_i , italic_x italic_x italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_k italic_q start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT + italic_c end_ARG - divide start_ARG 1 end_ARG start_ARG 4 end_ARG divide start_ARG italic_k start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT italic_q start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i , italic_x italic_x end_POSTSUBSCRIPT end_ARG start_ARG ( italic_k italic_q start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT + italic_c ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_ARG ) - end_CELL end_ROW start_ROW start_CELL - 2 โˆ‘ start_POSTSUBSCRIPT italic_j : italic_j โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( ( italic_k italic_q start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT + italic_c ) start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - 3 italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_k italic_q start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT + italic_c ) italic_q start_POSTSUBSCRIPT italic_j , italic_x italic_x end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - italic_k start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT italic_q start_POSTSUBSCRIPT italic_j , italic_x italic_x italic_x end_POSTSUBSCRIPT italic_ฮถ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (1.8)

where the short notation for partial derivatives is used: ฯ†ห™=โˆ‚tฯ†โข(t,x)ห™๐œ‘subscript๐‘ก๐œ‘๐‘ก๐‘ฅ{\dot{\varphi}}=\partial_{t}\varphi(t,x)overห™ start_ARG italic_ฯ† end_ARG = โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_ฯ† ( italic_t , italic_x ); ฯ†x=โˆ‚xฯ†โข(t,x)subscript๐œ‘๐‘ฅsubscript๐‘ฅ๐œ‘๐‘ก๐‘ฅ\varphi_{x}=\partial_{x}\varphi(t,x)italic_ฯ† start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT = โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_ฯ† ( italic_t , italic_x ), ฯ†xโขx=โˆ‚x2ฯ†โข(t,x)subscript๐œ‘๐‘ฅ๐‘ฅsubscriptsuperscript2๐‘ฅ๐œ‘๐‘ก๐‘ฅ\varphi_{xx}=\partial^{2}_{x}\varphi(t,x)italic_ฯ† start_POSTSUBSCRIPT italic_x italic_x end_POSTSUBSCRIPT = โˆ‚ start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_ฯ† ( italic_t , italic_x ) e.t.c., and ฮถโข(x)๐œ๐‘ฅ\zeta(x)italic_ฮถ ( italic_x ) is the Weierstrass ฮถ๐œ\zetaitalic_ฮถ-function. The parameter kโˆˆโ„‚๐‘˜โ„‚k\in\mathbb{C}italic_k โˆˆ blackboard_C stands behind โˆ‚xsubscript๐‘ฅ\partial_{x}โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT, and its power measures the number of x๐‘ฅxitalic_x-derivatives. The limit kโ†’0โ†’๐‘˜0k\rightarrow 0italic_k โ†’ 0 corresponds to the limit to finite-dimensional mechanics, when all the fields become independent of x๐‘ฅxitalic_x. In this limit equations (1.7)-(1.8) become the ordinary differential equations (1.1) with the normalization constant b~=โˆ’2โขc~๐‘2๐‘\tilde{b}=-2cover~ start_ARG italic_b end_ARG = - 2 italic_c.

In the field case the Lax equation (1.2) is replaced with the Zakharov-Shabat zero curvature equation [13, 6]:

โˆ‚tUโข(z)โˆ’kโขโˆ‚xVโข(z)+[Uโข(z),Vโข(z)]=0,Uโข(z),Vโข(z)โˆˆMatโข(N,โ„‚).formulae-sequencesubscript๐‘ก๐‘ˆ๐‘ง๐‘˜subscript๐‘ฅ๐‘‰๐‘ง๐‘ˆ๐‘ง๐‘‰๐‘ง0๐‘ˆ๐‘ง๐‘‰๐‘งMat๐‘โ„‚\begin{array}[]{c}\displaystyle{\partial_{t}{U}(z)-k\partial_{x}{V}(z)+[{U}(z)% ,{V}(z)]=0\,,\qquad{U}(z),{V}(z)\in{\rm Mat}(N,\mathbb{C})\,.}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_U ( italic_z ) - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_V ( italic_z ) + [ italic_U ( italic_z ) , italic_V ( italic_z ) ] = 0 , italic_U ( italic_z ) , italic_V ( italic_z ) โˆˆ roman_Mat ( italic_N , blackboard_C ) . end_CELL end_ROW end_ARRAY (1.9)

The infinite-dimensional analogue of the Liouville integrability is as follows. The integrals of motion come from the generating function trโข(Tkโข(z,2โขฯ€))trsuperscript๐‘‡๐‘˜๐‘ง2๐œ‹{\rm tr}(T^{k}(z,2\pi))roman_tr ( italic_T start_POSTSUPERSCRIPT italic_k end_POSTSUPERSCRIPT ( italic_z , 2 italic_ฯ€ ) ), where Tโข(z,x)=Pexpโข(1kโขโˆซ0x๐‘‘yโขUโข(z,y))๐‘‡๐‘ง๐‘ฅPexp1๐‘˜superscriptsubscript0๐‘ฅdifferential-d๐‘ฆ๐‘ˆ๐‘ง๐‘ฆT(z,x)={\rm Pexp}\Big{(}\frac{1}{k}\int\limits_{0}^{x}dy\,U(z,y)\Big{)}italic_T ( italic_z , italic_x ) = roman_Pexp ( divide start_ARG 1 end_ARG start_ARG italic_k end_ARG โˆซ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_x end_POSTSUPERSCRIPT italic_d italic_y italic_U ( italic_z , italic_y ) ) is the monodromy matrix. The involution property {trโข(Tkโข(z,2โขฯ€)),trโข(Tmโข(w,2โขฯ€))}=0trsuperscript๐‘‡๐‘˜๐‘ง2๐œ‹trsuperscript๐‘‡๐‘š๐‘ค2๐œ‹0\{{\rm tr}(T^{k}(z,2\pi)),{\rm tr}(T^{m}(w,2\pi))\}=0{ roman_tr ( italic_T start_POSTSUPERSCRIPT italic_k end_POSTSUPERSCRIPT ( italic_z , 2 italic_ฯ€ ) ) , roman_tr ( italic_T start_POSTSUPERSCRIPT italic_m end_POSTSUPERSCRIPT ( italic_w , 2 italic_ฯ€ ) ) } = 0 follows from existence of the classical r๐‘Ÿritalic_r-matrix structure of Maillet type. For the spinless 1+1 Calogero-Moser model (1.7)-(1.8) the existence of such r๐‘Ÿritalic_r-matrix structure was recently proved in [14].

The purpose of the paper is to describe the field analogue of the spin Calogero-Moser model (1.3)-(1.4). In fact, a part of this problem was already solved. The Lax matrix (the U๐‘ˆUitalic_U-matrix) for the field analogue of the spin Calogero-Moser model was derived in [10] using the group-theoretical approach to Hitchin systems on the affine Higgs bundles. It was also shown that the constraints ฮผi=constsubscript๐œ‡๐‘–const\mu_{i}={\rm const}italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = roman_const (1.5) underlying additional reduction are extended as ฮผiโข(x)=constsubscript๐œ‡๐‘–๐‘ฅconst\mu_{i}(x)={\rm const}italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = roman_const with

ฮผiโข(x)=๐’ฎiโขiโข(x)โˆ’kโขโˆ‚xqiโข(x),i=1,โ€ฆ,N,formulae-sequencesubscript๐œ‡๐‘–๐‘ฅsubscript๐’ฎ๐‘–๐‘–๐‘ฅ๐‘˜subscript๐‘ฅsubscript๐‘ž๐‘–๐‘ฅ๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{\mu_{i}(x)=\mathcal{S}_{ii}(x)-k\partial_{x}q_% {i}(x)\,,\quad i=1,...,N\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ( italic_x ) - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) , italic_i = 1 , โ€ฆ , italic_N , end_CELL end_ROW end_ARRAY (1.10)

where ๐’ฎiโขisubscript๐’ฎ๐‘–๐‘–\mathcal{S}_{ii}caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT are diagonal elements of the spin field matrix ๐’ฎ=๐’ฎโข(x)โˆˆMatโข(N,โ„‚)๐’ฎ๐’ฎ๐‘ฅMat๐‘โ„‚\mathcal{S}=\mathcal{S}(x)\in{\rm Mat}(N,\mathbb{C})caligraphic_S = caligraphic_S ( italic_x ) โˆˆ roman_Mat ( italic_N , blackboard_C ). These spin variables has the Poisson brackets {Siโขjโข(x),Skโขlโข(y)}=(โˆ’Siโขlโข(x)โขฮดkโขj+Skโขjโข(x)โขฮดiโขl)โขฮดโข(xโˆ’y)subscript๐‘†๐‘–๐‘—๐‘ฅsubscript๐‘†๐‘˜๐‘™๐‘ฆsubscript๐‘†๐‘–๐‘™๐‘ฅsubscript๐›ฟ๐‘˜๐‘—subscript๐‘†๐‘˜๐‘—๐‘ฅsubscript๐›ฟ๐‘–๐‘™๐›ฟ๐‘ฅ๐‘ฆ\{S_{ij}(x),S_{kl}(y)\}=(-S_{il}(x)\delta_{kj}+S_{kj}(x)\delta_{il})\delta(x-y){ italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_x ) , italic_S start_POSTSUBSCRIPT italic_k italic_l end_POSTSUBSCRIPT ( italic_y ) } = ( - italic_S start_POSTSUBSCRIPT italic_i italic_l end_POSTSUBSCRIPT ( italic_x ) italic_ฮด start_POSTSUBSCRIPT italic_k italic_j end_POSTSUBSCRIPT + italic_S start_POSTSUBSCRIPT italic_k italic_j end_POSTSUBSCRIPT ( italic_x ) italic_ฮด start_POSTSUBSCRIPT italic_i italic_l end_POSTSUBSCRIPT ) italic_ฮด ( italic_x - italic_y ) on the unreduced phase space โ€“ a coadjoint orbit of the loop group Lโข(GLN)๐ฟsubscriptGL๐‘L({\rm GL}_{N})italic_L ( roman_GL start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ) of ๐’žโˆžsuperscript๐’ž{\mathcal{C}}^{\infty}caligraphic_C start_POSTSUPERSCRIPT โˆž end_POSTSUPERSCRIPT-maps ๐•Š1โ†’GLNโ†’superscript๐•Š1subscriptGL๐‘{\mathbb{S}}^{1}\rightarrow{\rm GL}_{N}blackboard_S start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT โ†’ roman_GL start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT. This allows to describe the classical r๐‘Ÿritalic_r-matrix structure with unwanted term on the unreduced phase space [14].

In this paper the equations of motion are deduced. For this purpose we suggest an ansatz for V๐‘‰Vitalic_V-matrix satisfying the Zakharov-Shabat equation with (unwanted term) non-trivial right hand side:

โˆ‚tUโข(z)โˆ’kโขโˆ‚xVโข(z)+[Uโข(z),Vโข(z)]=unwantedโขterm.subscript๐‘ก๐‘ˆ๐‘ง๐‘˜subscript๐‘ฅ๐‘‰๐‘ง๐‘ˆ๐‘ง๐‘‰๐‘งunwantedterm\begin{array}[]{c}\displaystyle{\partial_{t}{U}(z)-k\partial_{x}{V}(z)+[{U}(z)% ,{V}(z)]={\rm unwanted\ term}\,.}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_U ( italic_z ) - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_V ( italic_z ) + [ italic_U ( italic_z ) , italic_V ( italic_z ) ] = roman_unwanted roman_term . end_CELL end_ROW end_ARRAY (1.11)

Similarly to the finite-dimensional case (1.5) the unwanted term vanishes on-shell the constraints ฮผiโข(x)=constsubscript๐œ‡๐‘–๐‘ฅconst\mu_{i}(x)={\rm const}italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = roman_const, thus providing integrable model on the reduced phase space. Our construction of U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair works in the special case, when the matrix of spin field variables ๐’ฎ๐’ฎ\mathcal{S}caligraphic_S satisfies the property

๐’ฎ2=c0โข๐’ฎsuperscript๐’ฎ2subscript๐‘0๐’ฎ\begin{array}[]{c}\displaystyle{\mathcal{S}^{2}=c_{0}\mathcal{S}}\end{array}start_ARRAY start_ROW start_CELL caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S end_CELL end_ROW end_ARRAY (1.12)

with some constant c0โˆˆโ„‚subscript๐‘0โ„‚c_{0}\in\mathbb{C}italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT โˆˆ blackboard_C, that is eigenvalues of ๐’ฎ๐’ฎ\mathcal{S}caligraphic_S are equal to either 00 or c0subscript๐‘0c_{0}italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT. The equations are as follows:

โˆ‚tpi+kโขโˆ‚xdi+โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโข(๐’ฎiโขi+๐’ฎmโขm)โขโ„˜โ€ฒโข(qiโขm)โˆ’โˆ‘mโ‰ iN(๐’ฎiโขmโขHmโขiโˆ’Hiโขmโข๐’ฎmโขi)โขโ„˜โข(qiโขm)=0subscript๐‘กsubscript๐‘๐‘–๐‘˜subscript๐‘ฅsubscript๐‘‘๐‘–superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘š๐‘šsuperscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘šsuperscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐ป๐‘š๐‘–subscript๐ป๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘š0\begin{array}[]{c}\displaystyle{\partial_{t}p_{i}+k\partial_{x}{d}_{i}+\sum% \limits_{m\neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}(\mathcal{S}_{ii}+% \mathcal{S}_{mm})\wp^{\prime}(q_{im})-\sum\limits_{m\neq i}^{N}\Big{(}\mathcal% {S}_{im}{H}_{mi}-{H}_{im}\mathcal{S}_{mi}\Big{)}\wp(q_{im})=0}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT italic_H start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT - italic_H start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) = 0 end_CELL end_ROW end_ARRAY (1.13)

and

โˆ‚t๐’ฎiโขjโˆ’kโขโˆ‚xHห˜iโขjโˆ’2โขc0โข(๐’ฎiโขiโˆ’๐’ฎjโขj)โข๐’ฎiโขjโขโ„˜โข(qiโขj)++(piโˆ’pj)Hiโขj+โˆ‘mโ‰ i,jN((๐’ฎiโขmHmโขjโˆ’Hiโขm๐’ฎmโขj)ฮท(qi,qm,qj)++โˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโข((๐’ฎiโขi+๐’ฎmโขm)โขโ„˜โข(qiโขm)โˆ’(๐’ฎjโขj+๐’ฎmโขm)โขโ„˜โข(qjโขm)+(๐’ฎiโขiโˆ’๐’ฎjโขj)โขโ„˜โข(qiโขj))=0\begin{array}[]{c}\displaystyle{\partial_{t}\mathcal{S}_{ij}-k\partial_{x}{% \breve{H}}_{ij}-2c_{0}(\mathcal{S}_{ii}-\mathcal{S}_{jj})\mathcal{S}_{ij}\wp(q% _{ij})+}\\ \\ \displaystyle{+(p_{i}-p_{j})H_{ij}+\sum\limits_{m\neq i,j}^{N}\Big{(}(\mathcal% {S}_{im}{H}_{mj}-{H}_{im}\mathcal{S}_{mj})\eta(q_{i},q_{m},q_{j})+}\\ \displaystyle{+\sum\limits_{m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}\Big% {(}(\mathcal{S}_{ii}+\mathcal{S}_{mm})\wp(q_{im})-(\mathcal{S}_{jj}+\mathcal{S% }_{mm})\wp(q_{jm})+(\mathcal{S}_{ii}-\mathcal{S}_{jj})\wp(q_{ij})\Big{)}=0}% \end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - 2 italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) italic_H start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT italic_H start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT - italic_H start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) + end_CELL end_ROW start_ROW start_CELL + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - ( caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j italic_m end_POSTSUBSCRIPT ) + ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) = 0 end_CELL end_ROW end_ARRAY (1.14)

for iโ‰ j๐‘–๐‘—i\neq jitalic_i โ‰  italic_j and i,j=1,โ€ฆ,Nformulae-sequence๐‘–๐‘—1โ€ฆ๐‘i,j=1,...,Nitalic_i , italic_j = 1 , โ€ฆ , italic_N, where the function ฮทโข(qi,qm,qj)๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\eta(q_{i},q_{m},q_{j})italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) is given in (3.8) and

H=โˆ’kc0โข[๐’ฎ,๐’ฎx]โˆˆMatโข(N,โ„‚),Hห˜=Hโˆ’OโˆˆMatโข(N,โ„‚),formulae-sequence๐ป๐‘˜subscript๐‘0๐’ฎsubscript๐’ฎ๐‘ฅMat๐‘โ„‚ห˜๐ป๐ป๐‘‚Mat๐‘โ„‚\begin{array}[]{c}\displaystyle{H=-\frac{k}{c_{0}}\,[\mathcal{S},\mathcal{S}_{% x}]\in{\rm Mat}(N,\mathbb{C})\,,\qquad{\breve{H}}=H-O\in{\rm Mat}(N,\mathbb{C}% )\,,}\end{array}start_ARRAY start_ROW start_CELL italic_H = - divide start_ARG italic_k end_ARG start_ARG italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT end_ARG [ caligraphic_S , caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ] โˆˆ roman_Mat ( italic_N , blackboard_C ) , overห˜ start_ARG italic_H end_ARG = italic_H - italic_O โˆˆ roman_Mat ( italic_N , blackboard_C ) , end_CELL end_ROW end_ARRAY (1.15)
Oiโขj=ฮดiโขjโข2โข(pi+ฮบ)โข๐’ฎiโขi+(1โˆ’ฮดiโขj)โข((pi+pj+2โขฮบ)โข๐’ฎiโขj+โˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโขฮทโข(qi,qm,qj)),subscript๐‘‚๐‘–๐‘—subscript๐›ฟ๐‘–๐‘—2subscript๐‘๐‘–๐œ…subscript๐’ฎ๐‘–๐‘–1subscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–subscript๐‘๐‘—2๐œ…subscript๐’ฎ๐‘–๐‘—superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{{O}_{ij}=\delta_{ij}2(p_{i}+\kappa)\mathcal{S}% _{ii}+(1-\delta_{ij})\Big{(}(p_{i}+p_{j}+2\kappa)\mathcal{S}_{ij}+\sum\limits_% {m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}\eta(q_{i},q_{m},q_{j})\Big{)}% \,,}\end{array}start_ARRAY start_ROW start_CELL italic_O start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT 2 ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT + 2 italic_ฮบ ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (1.16)
di=pi2+2โขฮบโขpiโˆ’โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โข(qiโขm),i=1,โ€ฆ,N,ฮบ=โˆ’1trโข(๐’ฎ)โขโˆ‘i=1N๐’ฎiโขiโขpi.formulae-sequencesubscript๐‘‘๐‘–superscriptsubscript๐‘๐‘–22๐œ…subscript๐‘๐‘–superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘šformulae-sequence๐‘–1โ€ฆ๐‘๐œ…1tr๐’ฎsuperscriptsubscript๐‘–1๐‘subscript๐’ฎ๐‘–๐‘–subscript๐‘๐‘–\begin{array}[]{c}\displaystyle{d_{i}=p_{i}^{2}+2\kappa p_{i}-\sum\limits_{m% \neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}\wp(q_{im})\,,\quad i=1,...,N\,,% \qquad\kappa=-\frac{1}{{\rm tr}(\mathcal{S})}\sum\limits_{i=1}^{N}\mathcal{S}_% {ii}p_{i}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + 2 italic_ฮบ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) , italic_i = 1 , โ€ฆ , italic_N , italic_ฮบ = - divide start_ARG 1 end_ARG start_ARG roman_tr ( caligraphic_S ) end_ARG โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (1.17)

Two requirements are fulfilled for the obtained equations of motion. On the one hand equations generalize the finite-dimensional case, that is (1.3)-(1.4) follow in the kโ†’0โ†’๐‘˜0k\rightarrow 0italic_k โ†’ 0 limit. On the other hand the spinless model (1.7)-(1.8) is reproduced in the special case rankโข(๐’ฎ)=1rank๐’ฎ1{\rm rank}(\mathcal{S})=1roman_rank ( caligraphic_S ) = 1.

Let us also mention that the field Calogero-Moser model is related to a certain class of Landau-Lifshitz models through IRF-Vertex relations [2], where the condition ๐’ฎ2=c0โข๐’ฎsuperscript๐’ฎ2subscript๐‘0๐’ฎ\mathcal{S}^{2}=c_{0}\mathcal{S}caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S plays the same important role [3]. The spinless field Calogero-Moser model can be also extended to the field analogue of the Ruijsenaars-Schneider model described in terms of semi-discrete Zakharov-Shabat equation [12].

The paper is organized as follows. In the next Section we recall the description of the classical finite-dimensional spin Calogero-Moser model and reduction to the spinless case. In Section 3 we briefly review the U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair of the spinless field Calogero-Moser model from [1]. In Section 4 we consider a natural ansatz for V๐‘‰Vitalic_V-matrix and derive the corresponding equations. However, a relation to the spinless model cannot be established at this stage. This problem is solved in Section 5, and the final answer for V๐‘‰Vitalic_V-matrix and equations of motion are obtained. Finally, it is shown in Section 6 how the obtained U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair is reduced to the one for the spinless field Calogero-Moser model. Definitions and properties of elliptic functions are given in the Appendix.

2 Spin Calogero-Moser model in classical mechanics

Unreduced phase space.

The classical spin Calogero-Moser model of glNsubscriptgl๐‘{\rm gl}_{N}roman_gl start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT type is described by the following Lax pair:

Liโขjspinโข(z)=ฮดiโขjโข(pi+SiโขiโขE1โข(z))+(1โˆ’ฮดiโขj)โขSiโขjโขฯ•โข(z,qiโˆ’qj),subscriptsuperscript๐ฟspin๐‘–๐‘—๐‘งsubscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–subscript๐‘†๐‘–๐‘–subscript๐ธ1๐‘ง1subscript๐›ฟ๐‘–๐‘—subscript๐‘†๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{L^{\hbox{\tiny{spin}}}_{ij}(z)=\delta_{ij}(p_{% i}+S_{ii}E_{1}(z))+(1-\delta_{ij})S_{ij}\phi(z,q_{i}-q_{j})\,,}\end{array}start_ARRAY start_ROW start_CELL italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) , end_CELL end_ROW end_ARRAY (2.1)
Miโขjspinโข(z)=โˆ’(1โˆ’ฮดiโขj)โขb~โขSiโขjโขfโข(z,qiโˆ’qj),subscriptsuperscript๐‘€spin๐‘–๐‘—๐‘ง1subscript๐›ฟ๐‘–๐‘—~๐‘subscript๐‘†๐‘–๐‘—๐‘“๐‘งsubscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{M^{\hbox{\tiny{spin}}}_{ij}(z)=-(1-\delta_{ij}% ){\tilde{b}}S_{ij}f(z,q_{i}-q_{j})\,,}\end{array}start_ARRAY start_ROW start_CELL italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = - ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) over~ start_ARG italic_b end_ARG italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) , end_CELL end_ROW end_ARRAY (2.2)

where the functions E1โข(z)subscript๐ธ1๐‘งE_{1}(z)italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ), ฯ•โข(z,u)italic-ฯ•๐‘ง๐‘ข\phi(z,u)italic_ฯ• ( italic_z , italic_u ) and fโข(z,u)๐‘“๐‘ง๐‘ขf(z,u)italic_f ( italic_z , italic_u ) are defined in the Appendix. This Lax pair satisfies the Lax equation with the unwanted term (1.5). Namely,

Lห™spinโข(z)+[Lspinโข(z),Mspinโข(z)]=โˆ’b~โขโˆ‘i,j=1NEiโขjโข(Siโขiโˆ’Sjโขj)โขSiโขjโขE1โข(z)โขfโข(z,qiโขj).superscriptห™๐ฟspin๐‘งsuperscript๐ฟspin๐‘งsuperscript๐‘€spin๐‘ง~๐‘superscriptsubscript๐‘–๐‘—1๐‘subscript๐ธ๐‘–๐‘—subscript๐‘†๐‘–๐‘–subscript๐‘†๐‘—๐‘—subscript๐‘†๐‘–๐‘—subscript๐ธ1๐‘ง๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{\dot{L}^{\hbox{\tiny{spin}}}(z)+[L^{\hbox{% \tiny{spin}}}(z),M^{\hbox{\tiny{spin}}}(z)]=-{\tilde{b}}\sum\limits_{i,j=1}^{N% }E_{ij}\,(S_{ii}-S_{jj})S_{ij}E_{1}(z)f(z,q_{ij})\,.}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_L end_ARG start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) + [ italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) , italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) ] = - over~ start_ARG italic_b end_ARG โˆ‘ start_POSTSUBSCRIPT italic_i , italic_j = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_E start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - italic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) . end_CELL end_ROW end_ARRAY (2.3)

The latter equation provides the equations of motion (1.3)-(1.4). The r.h.s. of (2.3) vanishes on the constraints Siโขi=constsubscript๐‘†๐‘–๐‘–constS_{ii}={\rm const}italic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = roman_const, i=1,โ€ฆ,N๐‘–1โ€ฆ๐‘i=1,...,Nitalic_i = 1 , โ€ฆ , italic_N. These are the first class constraints generated by the coadjoint action of the Cartan subgroup โ„‹โŠ‚GLNโ„‹subscriptGL๐‘{\mathcal{H}}\subset{\rm GL}_{N}caligraphic_H โŠ‚ roman_GL start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT on the coadjoint orbit ๐’ช๐’ช{\mathcal{O}}caligraphic_O corresponding to the spin matrix S๐‘†Sitalic_S (with fixed eigenvalues). Another N๐‘Nitalic_N constraints are some gauge fixation conditions โ€“ fixation of the action of โ„‹โ„‹{\mathcal{H}}caligraphic_H. Totally we have 2โขN2๐‘2N2 italic_N conditions, and these are the second class constraints. One can perform the Poisson reduction using the Dirac formula. The Dirac terms may change the Poisson brackets {Siโขj,Skโขl}subscript๐‘†๐‘–๐‘—subscript๐‘†๐‘˜๐‘™\{S_{ij},S_{kl}\}{ italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT , italic_S start_POSTSUBSCRIPT italic_k italic_l end_POSTSUBSCRIPT } and equations of motion. Notice that the final answer (for equations of motion) depends on the choice of the gauge fixation conditions, and there is no preferred way to choose them. This is why the classical spin Calogero-Moser model is usually described at the level of the unreduced phase space.

Spinless model.

Reduction to the spinless model (1.1) is performed as follows [4]. Consider a special case when the matrix of spin variables is of rank one:

Siโขj=ฮพiโขฯˆj.subscript๐‘†๐‘–๐‘—subscript๐œ‰๐‘–subscript๐œ“๐‘—\begin{array}[]{c}\displaystyle{S_{ij}=\xi_{i}\psi_{j}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (2.4)

Next, perform the gauge transformation

Lspinโข(z)โ†’gโˆ’1โขLspinโข(z)โขg,Mspinโข(z)โ†’gโˆ’1โขMspinโข(z)โขgโˆ’gโˆ’1โขgห™formulae-sequenceโ†’superscript๐ฟspin๐‘งsuperscript๐‘”1superscript๐ฟspin๐‘ง๐‘”โ†’superscript๐‘€spin๐‘งsuperscript๐‘”1superscript๐‘€spin๐‘ง๐‘”superscript๐‘”1ห™๐‘”\begin{array}[]{c}\displaystyle{L^{\hbox{\tiny{spin}}}(z)\rightarrow g^{-1}L^{% \hbox{\tiny{spin}}}(z)g\,,\qquad M^{\hbox{\tiny{spin}}}(z)\rightarrow g^{-1}M^% {\hbox{\tiny{spin}}}(z)g-g^{-1}{\dot{g}}}\end{array}start_ARRAY start_ROW start_CELL italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) โ†’ italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) italic_g , italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) โ†’ italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) italic_g - italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT overห™ start_ARG italic_g end_ARG end_CELL end_ROW end_ARRAY (2.5)

with

g=diagโข(ฮพ1,โ€ฆ,ฮพN)โˆˆMatโข(N,โ„‚).๐‘”diagsubscript๐œ‰1โ€ฆsubscript๐œ‰๐‘Mat๐‘โ„‚\begin{array}[]{c}\displaystyle{g={\rm diag}(\xi_{1},...,\xi_{N})\in{\rm Mat}(% N,\mathbb{C})\,.}\end{array}start_ARRAY start_ROW start_CELL italic_g = roman_diag ( italic_ฮพ start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , โ€ฆ , italic_ฮพ start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ) โˆˆ roman_Mat ( italic_N , blackboard_C ) . end_CELL end_ROW end_ARRAY (2.6)

The action on the matrix S๐‘†Sitalic_S is

Siโขjโ†’ฮพjโขฯˆj=Sjโขj,โ†’subscript๐‘†๐‘–๐‘—subscript๐œ‰๐‘—subscript๐œ“๐‘—subscript๐‘†๐‘—๐‘—\begin{array}[]{c}\displaystyle{S_{ij}\rightarrow\xi_{j}\psi_{j}=S_{jj}\,,}% \end{array}start_ARRAY start_ROW start_CELL italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT โ†’ italic_ฮพ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT = italic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT , end_CELL end_ROW end_ARRAY (2.7)

and Siโขi=constsubscript๐‘†๐‘–๐‘–constS_{ii}={\rm const}italic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = roman_const due to (1.4), that is the spin degrees of freedom are gauged away. Indeed, since the Lax matrix defines all Hamiltonian flows (and equations of motion) we conclude that the model is independent of spin variables in this case.

Due to (2.5) the M๐‘€Mitalic_M-matrix acquires additional diagonal part

โˆ’(gโˆ’1โขgห™)iโขj=โˆ’ฮดiโขjโขฮพห™iฮพi.subscriptsuperscript๐‘”1ห™๐‘”๐‘–๐‘—subscript๐›ฟ๐‘–๐‘—subscriptห™๐œ‰๐‘–subscript๐œ‰๐‘–\begin{array}[]{c}\displaystyle{-\Big{(}g^{-1}{\dot{g}}\Big{)}_{ij}=-\delta_{% ij}\frac{\dot{\xi}_{i}}{\xi_{i}}\,.}\end{array}start_ARRAY start_ROW start_CELL - ( italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT overห™ start_ARG italic_g end_ARG ) start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT divide start_ARG overห™ start_ARG italic_ฮพ end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG . end_CELL end_ROW end_ARRAY (2.8)

It can be computed through the Hamiltonian b~โขHspin~๐‘superscript๐ปspin{\tilde{b}}H^{\hbox{\tiny{spin}}}over~ start_ARG italic_b end_ARG italic_H start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT, where

Hspin=โˆ‘i=1Npi22โˆ’โˆ‘i>jNSiโขjโขSjโขiโขโ„˜โข(qiโˆ’qj),superscript๐ปspinsuperscriptsubscript๐‘–1๐‘superscriptsubscript๐‘๐‘–22superscriptsubscript๐‘–๐‘—๐‘subscript๐‘†๐‘–๐‘—subscript๐‘†๐‘—๐‘–Weierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{H^{\hbox{\tiny{spin}}}=\sum\limits_{i=1}^{N}% \frac{p_{i}^{2}}{2}-\sum\limits_{i>j}^{N}S_{ij}S_{ji}\wp(q_{i}-q_{j})\,,}\end{array}start_ARRAY start_ROW start_CELL italic_H start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT = โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT divide start_ARG italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_ARG start_ARG 2 end_ARG - โˆ‘ start_POSTSUBSCRIPT italic_i > italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_S start_POSTSUBSCRIPT italic_j italic_i end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) , end_CELL end_ROW end_ARRAY (2.9)

and the Poisson brackets {ฮพi,ฮทj}=ฮดiโขjsubscript๐œ‰๐‘–subscript๐œ‚๐‘—subscript๐›ฟ๐‘–๐‘—\{\xi_{i},\eta_{j}\}=\delta_{ij}{ italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_ฮท start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT } = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT. In this way the spinless model arise555Some more comments on the reduction to spinless model can be found in [4], see also [14].. The Lax pair of the spinless model [8]

LiโขjCMโข(z)=ฮดiโขjโข(pi+cโขE1โข(z))+cโข(1โˆ’ฮดiโขj)โขฯ•โข(z,qiโขj),qiโขj=qiโˆ’qj,formulae-sequencesubscriptsuperscript๐ฟCM๐‘–๐‘—๐‘งsubscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–๐‘subscript๐ธ1๐‘ง๐‘1subscript๐›ฟ๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐‘ž๐‘–๐‘—subscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{L^{\hbox{\tiny{CM}}}_{ij}(z)=\delta_{ij}(p_{i}% +{c}E_{1}(z))+{c}(1-\delta_{ij})\phi(z,q_{ij})\,,\quad q_{ij}=q_{i}-q_{j}\,,}% \end{array}start_ARRAY start_ROW start_CELL italic_L start_POSTSUPERSCRIPT CM end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_c italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) ) + italic_c ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT , end_CELL end_ROW end_ARRAY (2.10)
MiโขjCMโข(z)=โˆ’b~โขcโขฮดiโขjโขdiโˆ’b~โขcโข(1โˆ’ฮดiโขj)โขfโข(z,qiโขj),di=โˆ‘k:kโ‰ iNโ„˜โข(qiโขk)formulae-sequencesubscriptsuperscript๐‘€CM๐‘–๐‘—๐‘ง~๐‘๐‘subscript๐›ฟ๐‘–๐‘—subscript๐‘‘๐‘–~๐‘๐‘1subscript๐›ฟ๐‘–๐‘—๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐‘‘๐‘–superscriptsubscript:๐‘˜๐‘˜๐‘–๐‘Weierstrass-psubscript๐‘ž๐‘–๐‘˜\begin{array}[]{c}\displaystyle{M^{\hbox{\tiny{CM}}}_{ij}(z)=-{\tilde{b}}{c}% \delta_{ij}d_{i}-{\tilde{b}}{c}(1-\delta_{ij})f(z,q_{ij})\,,\quad d_{i}=\sum% \limits_{k:k\neq i}^{N}\wp(q_{ik})}\end{array}start_ARRAY start_ROW start_CELL italic_M start_POSTSUPERSCRIPT CM end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = - over~ start_ARG italic_b end_ARG italic_c italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - over~ start_ARG italic_b end_ARG italic_c ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) , italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = โˆ‘ start_POSTSUBSCRIPT italic_k : italic_k โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_k end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (2.11)

satisfies the Lax equation (1.2) and provides equations (1.1).

Alternatively, in order to deduce the spinless model one can use the equations of motion (1.4). In the case of rankโข(S)=1rank๐‘†1{\rm rank}(S)=1roman_rank ( italic_S ) = 1 (2.4) we have

Sห™iโขj=(ฮพห™iฮพi+ฯˆห™jฯˆj)โขฮพiโขฯˆj.subscriptห™๐‘†๐‘–๐‘—subscriptห™๐œ‰๐‘–subscript๐œ‰๐‘–subscriptห™๐œ“๐‘—subscript๐œ“๐‘—subscript๐œ‰๐‘–subscript๐œ“๐‘—\begin{array}[]{c}\displaystyle{{\dot{S}}_{ij}=\Big{(}\frac{\dot{\xi}_{i}}{\xi% _{i}}+\frac{\dot{\psi}_{j}}{\psi_{j}}\Big{)}\xi_{i}\psi_{j}\,.}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = ( divide start_ARG overห™ start_ARG italic_ฮพ end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG + divide start_ARG overห™ start_ARG italic_ฯˆ end_ARG start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT end_ARG start_ARG italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT end_ARG ) italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (2.12)

On the other hand from (1.4) we have

Sห™iโขj=b~โขฮพiโขฯˆjโขโˆ‘kโ‰ i,jNฮพkโขฯˆkโข(โ„˜โข(qiโˆ’qk)โˆ’โ„˜โข(qjโˆ’qk)),iโ‰ j.formulae-sequencesubscriptห™๐‘†๐‘–๐‘—~๐‘subscript๐œ‰๐‘–subscript๐œ“๐‘—superscriptsubscript๐‘˜๐‘–๐‘—๐‘subscript๐œ‰๐‘˜subscript๐œ“๐‘˜Weierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘˜Weierstrass-psubscript๐‘ž๐‘—subscript๐‘ž๐‘˜๐‘–๐‘—\begin{array}[]{c}\displaystyle{{\dot{S}}_{ij}={\tilde{b}}\xi_{i}\psi_{j}\sum% \limits_{k\neq i,j}^{N}\xi_{k}\psi_{k}(\wp(q_{i}-q_{k})-\wp(q_{j}-q_{k}))\,,\ % i\neq j\,.}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_k โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_ฮพ start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) ) , italic_i โ‰  italic_j . end_CELL end_ROW end_ARRAY (2.13)

The reduction consists of constraints (1.6) Siโขi=csubscript๐‘†๐‘–๐‘–๐‘S_{ii}=citalic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = italic_c and some gauge fixing conditions, which we choose for definiteness to be ฮพi=ฯˆi=c1/2subscript๐œ‰๐‘–subscript๐œ“๐‘–superscript๐‘12\xi_{i}=\psi_{i}=c^{1/2}italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_ฯˆ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_c start_POSTSUPERSCRIPT 1 / 2 end_POSTSUPERSCRIPT, i.e. on-shell the constraints we have Siโขj=csubscript๐‘†๐‘–๐‘—๐‘S_{ij}=citalic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_c, and from (2.13) one gets

Sห™iโขj|onโˆ’shell=b~โขc2โขโˆ‘kโ‰ i,jN(โ„˜โข(qiโˆ’qk)โˆ’โ„˜โข(qjโˆ’qk)),iโ‰ j.formulae-sequenceevaluated-atsubscriptห™๐‘†๐‘–๐‘—onshell~๐‘superscript๐‘2superscriptsubscript๐‘˜๐‘–๐‘—๐‘Weierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘˜Weierstrass-psubscript๐‘ž๐‘—subscript๐‘ž๐‘˜๐‘–๐‘—\begin{array}[]{c}\displaystyle{{\dot{S}}_{ij}\Big{|}_{\rm on-shell}={\tilde{b% }}c^{2}\sum\limits_{k\neq i,j}^{N}(\wp(q_{i}-q_{k})-\wp(q_{j}-q_{k}))\,,\ i% \neq j\,.}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT | start_POSTSUBSCRIPT roman_on - roman_shell end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_k โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT ) ) , italic_i โ‰  italic_j . end_CELL end_ROW end_ARRAY (2.14)

Therefore, from (2.12) we conclude that

ฮพห™iฮพi|onโˆ’shell=b~โขc2โขโˆ‘mโ‰ iNโ„˜โข(qiโˆ’qm)+ฮปห™,evaluated-atsubscriptห™๐œ‰๐‘–subscript๐œ‰๐‘–onshell~๐‘superscript๐‘2superscriptsubscript๐‘š๐‘–๐‘Weierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘šห™๐œ†\begin{array}[]{c}\displaystyle{\frac{\dot{\xi}_{i}}{\xi_{i}}\Big{|}_{\rm on-% shell}={\tilde{b}}c^{2}\sum\limits_{m\neq i}^{N}\wp(q_{i}-q_{m})+{\dot{\lambda% }}\,,}\end{array}start_ARRAY start_ROW start_CELL divide start_ARG overห™ start_ARG italic_ฮพ end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG | start_POSTSUBSCRIPT roman_on - roman_shell end_POSTSUBSCRIPT = over~ start_ARG italic_b end_ARG italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT ) + overห™ start_ARG italic_ฮป end_ARG , end_CELL end_ROW end_ARRAY (2.15)

where ฮปโข(t)๐œ†๐‘ก\lambda(t)italic_ฮป ( italic_t ) is some function coming from a freedom in definition of ฮพisubscript๐œ‰๐‘–\xi_{i}italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, ฯˆjsubscript๐œ“๐‘—\psi_{j}italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT: ฮพiโ†’eฮปโขฮพiโ†’subscript๐œ‰๐‘–superscript๐‘’๐œ†subscript๐œ‰๐‘–\xi_{i}\rightarrow e^{\lambda}\xi_{i}italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT โ†’ italic_e start_POSTSUPERSCRIPT italic_ฮป end_POSTSUPERSCRIPT italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, ฯˆjโ†’eโˆ’ฮปโขฯˆjโ†’subscript๐œ“๐‘—superscript๐‘’๐œ†subscript๐œ“๐‘—\psi_{j}\rightarrow e^{-\lambda}\psi_{j}italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT โ†’ italic_e start_POSTSUPERSCRIPT - italic_ฮป end_POSTSUPERSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT. Finally, plugging (2.15), (2.8) into (2.5) yields

(gโˆ’1โขMspinโข(z)โขgโˆ’gโˆ’1โขgห™)|onโˆ’shell=MCMโข(z)โˆ’1Nโขฮปห™evaluated-atsuperscript๐‘”1superscript๐‘€spin๐‘ง๐‘”superscript๐‘”1ห™๐‘”onshellsuperscript๐‘€CM๐‘งsubscript1๐‘ห™๐œ†\begin{array}[]{c}\displaystyle{\Big{(}g^{-1}M^{\hbox{\tiny{spin}}}(z)g-g^{-1}% {\dot{g}}\Big{)}\Big{|}_{\rm on-shell}=M^{\hbox{\tiny{CM}}}(z)-1_{N}{\dot{% \lambda}}}\end{array}start_ARRAY start_ROW start_CELL ( italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) italic_g - italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT overห™ start_ARG italic_g end_ARG ) | start_POSTSUBSCRIPT roman_on - roman_shell end_POSTSUBSCRIPT = italic_M start_POSTSUPERSCRIPT CM end_POSTSUPERSCRIPT ( italic_z ) - 1 start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT overห™ start_ARG italic_ฮป end_ARG end_CELL end_ROW end_ARRAY (2.16)

The term 1Nโขฮปห™subscript1๐‘ห™๐œ†1_{N}{\dot{\lambda}}1 start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT overห™ start_ARG italic_ฮป end_ARG does not provide any input into the Lax equation, and it can be gauged away.

3 Spinless field Calogero-Moser model

Here we recall main result of [1]. Below we use slightly different notations, normalization constants and gauge choice for U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair, see [14]. From now on qi=qiโข(t,x)subscript๐‘ž๐‘–subscript๐‘ž๐‘–๐‘ก๐‘ฅq_{i}=q_{i}(t,x)italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_t , italic_x ), pi=piโข(t,x)subscript๐‘๐‘–subscript๐‘๐‘–๐‘ก๐‘ฅp_{i}=p_{i}(t,x)italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_t , italic_x ), i=1,โ€ฆ,N๐‘–1โ€ฆ๐‘i=1,...,Nitalic_i = 1 , โ€ฆ , italic_N and ๐’ฎiโขj=๐’ฎiโขjโข(x)subscript๐’ฎ๐‘–๐‘—subscript๐’ฎ๐‘–๐‘—๐‘ฅ\mathcal{S}_{ij}=\mathcal{S}_{ij}(x)caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_x ), i,j=1,โ€ฆ,Nformulae-sequence๐‘–๐‘—1โ€ฆ๐‘i,j=1,...,Nitalic_i , italic_j = 1 , โ€ฆ , italic_N are periodic fields on a circle, and the time dependence comes from equations of motion generated by the Zakharov-Shabat equation.

Introduce notations

ฮฑi=(kโขqi,x+c)1/2,i=1,โ€ฆ,Nformulae-sequencesubscript๐›ผ๐‘–superscript๐‘˜subscript๐‘ž๐‘–๐‘ฅ๐‘12๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{\alpha_{i}=(kq_{i,x}+c)^{1/2}\,,\quad i=1,...,% N}\end{array}start_ARRAY start_ROW start_CELL italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = ( italic_k italic_q start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT + italic_c ) start_POSTSUPERSCRIPT 1 / 2 end_POSTSUPERSCRIPT , italic_i = 1 , โ€ฆ , italic_N end_CELL end_ROW end_ARRAY (3.1)

and

ฮบ=โˆ’1Nโขcโขโˆ‘j=1Npjโข(kโขqj,x+c)=โˆ’1Nโขcโขโˆ‘j=1Npjโขฮฑj2.๐œ…1๐‘๐‘superscriptsubscript๐‘—1๐‘subscript๐‘๐‘—๐‘˜subscript๐‘ž๐‘—๐‘ฅ๐‘1๐‘๐‘superscriptsubscript๐‘—1๐‘subscript๐‘๐‘—superscriptsubscript๐›ผ๐‘—2\begin{array}[]{c}\displaystyle{\kappa=-\frac{1}{Nc}\sum\limits_{j=1}^{N}p_{j}% (kq_{j,x}+c)=-\frac{1}{Nc}\sum\limits_{j=1}^{N}p_{j}\alpha_{j}^{2}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_ฮบ = - divide start_ARG 1 end_ARG start_ARG italic_N italic_c end_ARG โˆ‘ start_POSTSUBSCRIPT italic_j = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_k italic_q start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT + italic_c ) = - divide start_ARG 1 end_ARG start_ARG italic_N italic_c end_ARG โˆ‘ start_POSTSUBSCRIPT italic_j = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT . end_CELL end_ROW end_ARRAY (3.2)

In what follows we also use the short notation

qiโขj=qiโˆ’qjsubscript๐‘ž๐‘–๐‘—subscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{q_{ij}=q_{i}-q_{j}}\end{array}start_ARRAY start_ROW start_CELL italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (3.3)

as in (2.10)-(2.11). Consider the U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair

Uiโขj2dCMโข(z)=ฮดiโขjโข(pi+ฮฑi2โขE1โข(z)โˆ’kโขฮฑi,xฮฑi)+(1โˆ’ฮดiโขj)โขฯ•โข(z,qiโˆ’qj)โขฮฑj2,subscriptsuperscript๐‘ˆ2dCM๐‘–๐‘—๐‘งsubscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–superscriptsubscript๐›ผ๐‘–2subscript๐ธ1๐‘ง๐‘˜subscript๐›ผ๐‘–๐‘ฅsubscript๐›ผ๐‘–1subscript๐›ฟ๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–subscript๐‘ž๐‘—superscriptsubscript๐›ผ๐‘—2\begin{array}[]{c}\displaystyle{U^{\hbox{\tiny{2dCM}}}_{ij}(z)=\delta_{ij}\Big% {(}p_{i}+\alpha_{i}^{2}E_{1}(z)-k\frac{\alpha_{i,x}}{\alpha_{i}}\Big{)}+(1-% \delta_{ij})\phi(z,q_{i}-q_{j})\alpha_{j}^{2}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_U start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) - italic_k divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , end_CELL end_ROW end_ARRAY (3.4)
Viโขj2dCMโข(z)=ฮดiโขjโข(qi,tโขE1โข(z)โˆ’Nโขcโขฮฑi2โขโ„˜โข(z)โˆ’mi0โˆ’ฮฑi,tฮฑi)++(1โˆ’ฮดiโขj)โข(Nโขcโขfโข(z,qiโˆ’qj)โˆ’NโขcโขE1โข(z)โขฯ•โข(z,qiโˆ’qj)โˆ’miโขjโขฯ•โข(z,qiโˆ’qj))โขฮฑj2,subscriptsuperscript๐‘‰2dCM๐‘–๐‘—๐‘งlimit-fromsubscript๐›ฟ๐‘–๐‘—subscript๐‘ž๐‘–๐‘กsubscript๐ธ1๐‘ง๐‘๐‘superscriptsubscript๐›ผ๐‘–2Weierstrass-p๐‘งsuperscriptsubscript๐‘š๐‘–0subscript๐›ผ๐‘–๐‘กsubscript๐›ผ๐‘–missing-subexpression1subscript๐›ฟ๐‘–๐‘—๐‘๐‘๐‘“๐‘งsubscript๐‘ž๐‘–subscript๐‘ž๐‘—๐‘๐‘subscript๐ธ1๐‘งitalic-ฯ•๐‘งsubscript๐‘ž๐‘–subscript๐‘ž๐‘—subscript๐‘š๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–subscript๐‘ž๐‘—superscriptsubscript๐›ผ๐‘—2\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dCM}}}_{ij}(z)=\delta_{ij}\Big% {(}q_{i,t}E_{1}(z)-Nc\alpha_{i}^{2}\wp(z)-m_{i}^{0}-\frac{\alpha_{i,t}}{\alpha% _{i}}\Big{)}+}\\ \\ \displaystyle{+(1-\delta_{ij})\Big{(}Ncf(z,q_{i}-q_{j})-NcE_{1}(z)\phi(z,q_{i}% -q_{j})-m_{ij}\phi(z,q_{i}-q_{j})\Big{)}\alpha_{j}^{2}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i , italic_t end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) - italic_N italic_c italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT โ„˜ ( italic_z ) - italic_m start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT - divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_t end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG ) + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( italic_N italic_c italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - italic_N italic_c italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - italic_m start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) ) italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , end_CELL end_ROW end_ARRAY (3.5)

where

mi0=pi2+2โขฮบโขpi+k2โขฮฑi,xโขxฮฑiโˆ’โˆ‘j:jโ‰ iN((2โขฮฑj4+ฮฑi2โขฮฑj2)โขโ„˜โข(qiโˆ’qj)+4โขkโขฮฑjโขฮฑj,xโขฮถโข(qiโˆ’qj)),superscriptsubscript๐‘š๐‘–0superscriptsubscript๐‘๐‘–22๐œ…subscript๐‘๐‘–superscript๐‘˜2subscript๐›ผ๐‘–๐‘ฅ๐‘ฅsubscript๐›ผ๐‘–superscriptsubscript:๐‘—๐‘—๐‘–๐‘2superscriptsubscript๐›ผ๐‘—4superscriptsubscript๐›ผ๐‘–2superscriptsubscript๐›ผ๐‘—2Weierstrass-psubscript๐‘ž๐‘–subscript๐‘ž๐‘—4๐‘˜subscript๐›ผ๐‘—subscript๐›ผ๐‘—๐‘ฅ๐œsubscript๐‘ž๐‘–subscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{m_{i}^{0}=p_{i}^{2}+2\kappa p_{i}+k^{2}\frac{% \alpha_{i,xx}}{\alpha_{i}}-\sum\limits_{j:j\neq i}^{N}\Big{(}(2\alpha_{j}^{4}+% \alpha_{i}^{2}\alpha_{j}^{2})\wp(q_{i}-q_{j})+4k\alpha_{j}\alpha_{j,x}\zeta(q_% {i}-q_{j})\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_m start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT = italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + 2 italic_ฮบ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG - โˆ‘ start_POSTSUBSCRIPT italic_j : italic_j โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( ( 2 italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT + italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) + 4 italic_k italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT italic_ฮถ ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (3.6)

and

miโขj=pi+pj+2โขฮบโˆ’kโขฮฑi,xฮฑi+kโขฮฑj,xฮฑj+โˆ‘l:lโ‰ i,jNฮฑl2โขฮทโข(qi,ql,qj),iโ‰ jformulae-sequencesubscript๐‘š๐‘–๐‘—subscript๐‘๐‘–subscript๐‘๐‘—2๐œ…๐‘˜subscript๐›ผ๐‘–๐‘ฅsubscript๐›ผ๐‘–๐‘˜subscript๐›ผ๐‘—๐‘ฅsubscript๐›ผ๐‘—superscriptsubscript:๐‘™๐‘™๐‘–๐‘—๐‘superscriptsubscript๐›ผ๐‘™2๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘™subscript๐‘ž๐‘—๐‘–๐‘—\begin{array}[]{c}\displaystyle{m_{ij}=p_{i}+p_{j}+2\kappa-k\frac{\alpha_{i,x}% }{\alpha_{i}}+k\frac{\alpha_{j,x}}{\alpha_{j}}+\sum\limits_{l:l\neq i,j}^{N}% \alpha_{l}^{2}\eta(q_{i},q_{l},q_{j})\,,\quad i\neq j}\end{array}start_ARRAY start_ROW start_CELL italic_m start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT + 2 italic_ฮบ - italic_k divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG + italic_k divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT end_ARG + โˆ‘ start_POSTSUBSCRIPT italic_l : italic_l โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_l end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_l end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) , italic_i โ‰  italic_j end_CELL end_ROW end_ARRAY (3.7)

with

ฮทโข(z1,z2,z3)=ฮถโข(z1โˆ’z2)+ฮถโข(z2โˆ’z3)+ฮถโข(z3โˆ’z1)=(A.5)=E1โข(z1โˆ’z2)+E1โข(z2โˆ’z3)+E1โข(z3โˆ’z1).๐œ‚subscript๐‘ง1subscript๐‘ง2subscript๐‘ง3๐œsubscript๐‘ง1subscript๐‘ง2๐œsubscript๐‘ง2subscript๐‘ง3๐œsubscript๐‘ง3subscript๐‘ง1superscriptA.5absentmissing-subexpressionabsentsubscript๐ธ1subscript๐‘ง1subscript๐‘ง2subscript๐ธ1subscript๐‘ง2subscript๐‘ง3subscript๐ธ1subscript๐‘ง3subscript๐‘ง1\begin{array}[]{c}\displaystyle{\eta(z_{1},z_{2},z_{3})=\zeta(z_{1}-z_{2})+% \zeta(z_{2}-z_{3})+\zeta(z_{3}-z_{1})\stackrel{{\scriptstyle(\ref{a05})}}{{=}}% }\\ \\ \displaystyle{=E_{1}(z_{1}-z_{2})+E_{1}(z_{2}-z_{3})+E_{1}(z_{3}-z_{1})\,.}% \end{array}start_ARRAY start_ROW start_CELL italic_ฮท ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT , italic_z start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT ) = italic_ฮถ ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) + italic_ฮถ ( italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT ) + italic_ฮถ ( italic_z start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) start_RELOP SUPERSCRIPTOP start_ARG = end_ARG start_ARG ( ) end_ARG end_RELOP end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL = italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) + italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT ) + italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) . end_CELL end_ROW end_ARRAY (3.8)

Then the Zakharov-Shabat equation (1.9) for (3.4)-(3.5) provides the equations of motion (1.7)-(1.8), which are rewritten in the notations (3.1)-(3.2) as

qห™i=โˆ’2โขฮฑi2โข(pi+ฮบ),subscriptห™๐‘ž๐‘–2superscriptsubscript๐›ผ๐‘–2subscript๐‘๐‘–๐œ…\begin{array}[]{c}\displaystyle{{\dot{q}}_{i}=-2\alpha_{i}^{2}(p_{i}+\kappa)\,% ,}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = - 2 italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ ) , end_CELL end_ROW end_ARRAY (3.9)
pห™i=โˆ’kโขโˆ‚x(pi2+2โขฮบโขpi+k2โขฮฑi,xโขxฮฑi)โˆ’2โขโˆ‘j:jโ‰ iN(ฮฑj6โขโ„˜โ€ฒโข(qiโขj)โˆ’6โขฮฑj3โขฮฑj,xโขโ„˜โข(qiโขj)โˆ’k2โขโˆ‚x2(ฮฑj2)โขฮถโข(qiโขj)).subscriptห™๐‘๐‘–๐‘˜subscript๐‘ฅsuperscriptsubscript๐‘๐‘–22๐œ…subscript๐‘๐‘–superscript๐‘˜2subscript๐›ผ๐‘–๐‘ฅ๐‘ฅsubscript๐›ผ๐‘–2superscriptsubscript:๐‘—๐‘—๐‘–๐‘superscriptsubscript๐›ผ๐‘—6superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘—6superscriptsubscript๐›ผ๐‘—3subscript๐›ผ๐‘—๐‘ฅWeierstrass-psubscript๐‘ž๐‘–๐‘—superscript๐‘˜2superscriptsubscript๐‘ฅ2superscriptsubscript๐›ผ๐‘—2๐œsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{{\dot{p}}_{i}=-k\partial_{x}\Big{(}p_{i}^{2}+2% \kappa p_{i}+k^{2}\frac{\alpha_{i,xx}}{\alpha_{i}}\Big{)}-2\sum\limits_{j:j% \neq i}^{N}\Big{(}\alpha_{j}^{6}\wp^{\prime}(q_{ij})-6\alpha_{j}^{3}\alpha_{j,% x}\wp(q_{ij})-k^{2}\partial_{x}^{2}(\alpha_{j}^{2})\zeta(q_{ij})\Big{)}\,.}% \end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_p end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + 2 italic_ฮบ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG ) - 2 โˆ‘ start_POSTSUBSCRIPT italic_j : italic_j โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 6 end_POSTSUPERSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) - 6 italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) - italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) italic_ฮถ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) . end_CELL end_ROW end_ARRAY (3.10)

These equations are Hamiltonian with the Hamiltonian function

โ„‹2dCM=โˆซ02โขฯ€๐‘‘xโขH2dCMโข(x),H2dCMโข(x)=โˆ’โˆ‘i=1Npi2โขฮฑi2+Nโขcโขฮบ2+โˆ‘i=1Nk2โขฮฑi,x2++k2โขโˆ‘iโ‰ jN(ฮฑiโขฮฑj,xโˆ’ฮฑjโขฮฑi,x+cโข(ฮฑi,xโˆ’ฮฑj,x))โขฮถโข(qiโขj)+12โขโˆ‘iโ‰ jN(ฮฑi4โขฮฑj2+ฮฑi2โขฮฑj4โˆ’cโข(ฮฑi2โˆ’ฮฑj2)2)โขโ„˜โข(qiโขj).formulae-sequencesuperscriptโ„‹2dCMsuperscriptsubscript02๐œ‹differential-d๐‘ฅsuperscript๐ป2dCM๐‘ฅsuperscript๐ป2dCM๐‘ฅsuperscriptsubscript๐‘–1๐‘superscriptsubscript๐‘๐‘–2superscriptsubscript๐›ผ๐‘–2๐‘๐‘superscript๐œ…2limit-fromsuperscriptsubscript๐‘–1๐‘superscript๐‘˜2superscriptsubscript๐›ผ๐‘–๐‘ฅ2๐‘˜2superscriptsubscript๐‘–๐‘—๐‘subscript๐›ผ๐‘–subscript๐›ผ๐‘—๐‘ฅsubscript๐›ผ๐‘—subscript๐›ผ๐‘–๐‘ฅ๐‘subscript๐›ผ๐‘–๐‘ฅsubscript๐›ผ๐‘—๐‘ฅ๐œsubscript๐‘ž๐‘–๐‘—12superscriptsubscript๐‘–๐‘—๐‘superscriptsubscript๐›ผ๐‘–4superscriptsubscript๐›ผ๐‘—2superscriptsubscript๐›ผ๐‘–2superscriptsubscript๐›ผ๐‘—4๐‘superscriptsuperscriptsubscript๐›ผ๐‘–2superscriptsubscript๐›ผ๐‘—22Weierstrass-psubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{{\mathcal{H}}^{\hbox{\tiny{2dCM}}}=\int\limits% _{0}^{2\pi}dxH^{\hbox{\tiny{2dCM}}}(x)\,,\qquad H^{\hbox{\tiny{2dCM}}}(x)=-% \sum\limits_{i=1}^{N}p_{i}^{2}\alpha_{i}^{2}+Nc\kappa^{2}+\sum\limits_{i=1}^{N% }k^{2}\alpha_{i,x}^{2}+}\\ \displaystyle{+\frac{k}{2}\sum\limits_{i\neq j}^{N}\Big{(}\alpha_{i}\alpha_{j,% x}-\alpha_{j}\alpha_{i,x}+c(\alpha_{i,x}-\alpha_{j,x})\Big{)}\zeta(q_{ij})+% \frac{1}{2}\sum\limits_{i\neq j}^{N}\Big{(}\alpha_{i}^{4}\alpha_{j}^{2}+\alpha% _{i}^{2}\alpha_{j}^{4}-c(\alpha_{i}^{2}-\alpha_{j}^{2})^{2}\Big{)}\wp(q_{ij})% \,.}\end{array}start_ARRAY start_ROW start_CELL caligraphic_H start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT = โˆซ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 italic_ฯ€ end_POSTSUPERSCRIPT italic_d italic_x italic_H start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT ( italic_x ) , italic_H start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT ( italic_x ) = - โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + italic_N italic_c italic_ฮบ start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + end_CELL end_ROW start_ROW start_CELL + divide start_ARG italic_k end_ARG start_ARG 2 end_ARG โˆ‘ start_POSTSUBSCRIPT italic_i โ‰  italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT - italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT + italic_c ( italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT - italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT ) ) italic_ฮถ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + divide start_ARG 1 end_ARG start_ARG 2 end_ARG โˆ‘ start_POSTSUBSCRIPT italic_i โ‰  italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT - italic_c ( italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT - italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) . end_CELL end_ROW end_ARRAY (3.11)

In the limit to the finite-dimensional case k=0๐‘˜0k=0italic_k = 0. Then ฮบ=1Nโขโˆ‘m=1Npm๐œ…1๐‘superscriptsubscript๐‘š1๐‘subscript๐‘๐‘š\kappa=\frac{1}{N}\sum\limits_{m=1}^{N}p_{m}italic_ฮบ = divide start_ARG 1 end_ARG start_ARG italic_N end_ARG โˆ‘ start_POSTSUBSCRIPT italic_m = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT, ฮฑi2=csuperscriptsubscript๐›ผ๐‘–2๐‘\alpha_{i}^{2}=citalic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c, and we come to (1.1) with b~=โˆ’2โขc~๐‘2๐‘{\tilde{b}}=-2cover~ start_ARG italic_b end_ARG = - 2 italic_c.

4 1+1 field spin Calogero-Moser model: ansatz for U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair

The U๐‘ˆUitalic_U-matrix for the 1+1 field spin Calogero-Moser model was derived in [10] by extending the corresponding moment map equation (defining relation for the Lax matrix) to the case of loop group. The result is

Uiโขj2dSpinโข(z)=ฮดiโขjโข(pi+๐’ฎiโขiโข(x)โขE1โข(z))+(1โˆ’ฮดiโขj)โข๐’ฎiโขjโข(x)โขฯ•โข(z,qiโข(x)โˆ’qjโข(x)),subscriptsuperscript๐‘ˆ2dSpin๐‘–๐‘—๐‘งsubscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–subscript๐’ฎ๐‘–๐‘–๐‘ฅsubscript๐ธ1๐‘ง1subscript๐›ฟ๐‘–๐‘—subscript๐’ฎ๐‘–๐‘—๐‘ฅitalic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘ฅsubscript๐‘ž๐‘—๐‘ฅ\begin{array}[]{c}\displaystyle{U^{\hbox{\tiny{2dSpin}}}_{ij}(z)=\delta_{ij}(p% _{i}+\mathcal{S}_{ii}(x)E_{1}(z))+(1-\delta_{ij})\mathcal{S}_{ij}(x)\phi(z,q_{% i}(x)-q_{j}(x))\,,}\end{array}start_ARRAY start_ROW start_CELL italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ( italic_x ) italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_x ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) - italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ( italic_x ) ) , end_CELL end_ROW end_ARRAY (4.1)

and it has exactly the same form as in the finite dimensional case (2.1) but the additional constraint is different. It is now

ฮผiโข(x)=c,i=1,โ€ฆ,Nformulae-sequencesubscript๐œ‡๐‘–๐‘ฅ๐‘๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{\mu_{i}(x)=c\,,\quad i=1,...,N}\end{array}start_ARRAY start_ROW start_CELL italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = italic_c , italic_i = 1 , โ€ฆ , italic_N end_CELL end_ROW end_ARRAY (4.2)

where ฮผiโข(x)subscript๐œ‡๐‘–๐‘ฅ\mu_{i}(x)italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) is (1.10) and cโˆˆโ„‚๐‘โ„‚c\in\mathbb{C}italic_c โˆˆ blackboard_C is some constant. Equivalently,

๐’ฎiโขiโข(x)โˆ’kโขโˆ‚xqiโข(x)=cor๐’ฎiโขiโข(x)=ฮฑi2โข(x),i=1,โ€ฆ,N.formulae-sequencesubscript๐’ฎ๐‘–๐‘–๐‘ฅ๐‘˜subscript๐‘ฅsubscript๐‘ž๐‘–๐‘ฅ๐‘orformulae-sequencesubscript๐’ฎ๐‘–๐‘–๐‘ฅsuperscriptsubscript๐›ผ๐‘–2๐‘ฅ๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{\mathcal{S}_{ii}(x)-k\partial_{x}q_{i}(x)=c% \quad\hbox{or}\quad\mathcal{S}_{ii}(x)=\alpha_{i}^{2}(x)\,,\quad i=1,...,N\,.}% \end{array}start_ARRAY start_ROW start_CELL caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ( italic_x ) - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = italic_c or caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ( italic_x ) = italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_x ) , italic_i = 1 , โ€ฆ , italic_N . end_CELL end_ROW end_ARRAY (4.3)

However, the group-theoretical approach of [10] does not provide any constructive approach for explicit calculation of V๐‘‰Vitalic_V-matrix. Here we use an ansatz, which is similar to the one for glNsubscriptgl๐‘{\rm gl}_{N}roman_gl start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT Landau-Lifshitz equation [15, 3]. Consider the following V๐‘‰Vitalic_V-matrix:

Viโขjโข(z)=ฮดiโขjโข(d~i+H~iโขiโขE1โข(z)+bโข๐’ฎiโขiโขโ„˜โข(z))+(1โˆ’ฮดiโขj)โข(H~iโขjโขฯ•โข(z,qiโขj)+bโข๐’ฎiโขjโขFโข(z,qiโขj)),subscript๐‘‰๐‘–๐‘—๐‘งsubscript๐›ฟ๐‘–๐‘—subscript~๐‘‘๐‘–subscript~๐ป๐‘–๐‘–subscript๐ธ1๐‘ง๐‘subscript๐’ฎ๐‘–๐‘–Weierstrass-p๐‘ง1subscript๐›ฟ๐‘–๐‘—subscript~๐ป๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘—๐น๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{V_{ij}(z)=\delta_{ij}\Big{(}{\tilde{d}}_{i}+{% \tilde{H}}_{ii}E_{1}(z)+b\mathcal{S}_{ii}\wp(z)\Big{)}+(1-\delta_{ij})\Big{(}{% \tilde{H}}_{ij}\phi(z,q_{ij})+b\mathcal{S}_{ij}F(z,q_{ij})\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + italic_b caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT โ„˜ ( italic_z ) ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + italic_b caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (4.4)

where the function Fโข(z,q)๐น๐‘ง๐‘žF(z,q)italic_F ( italic_z , italic_q ) is given by (A.13) and H~=H~โข(t,x)โˆˆMatโข(N,โ„‚)~๐ป~๐ป๐‘ก๐‘ฅMat๐‘โ„‚{\tilde{H}}={\tilde{H}}(t,x)\in{\rm Mat}(N,\mathbb{C})over~ start_ARG italic_H end_ARG = over~ start_ARG italic_H end_ARG ( italic_t , italic_x ) โˆˆ roman_Mat ( italic_N , blackboard_C ) is some matrix to be defined.

Theorem 1

The U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair (4.1) and (4.4) satisfies the Zakharov-Shabat equation with the unwanted term

โˆ‚tU2dSpinโข(z)โˆ’kโขโˆ‚xVโข(z)+[U2dSpinโข(z),Vโข(z)]==โˆ‘iโ‰ jNEiโขjโข(bโข๐’ฎiโขjโขFโ€ฒโข(z,qiโขj)+H~iโขjโขfโข(z,qiโขj))โข(ฮผiโข(x)โˆ’ฮผiโข(x)),ฮผiโข(x)=๐’ฎiโขiโˆ’kโขโˆ‚xqisubscript๐‘กsuperscript๐‘ˆ2dSpin๐‘ง๐‘˜subscript๐‘ฅ๐‘‰๐‘งsuperscript๐‘ˆ2dSpin๐‘ง๐‘‰๐‘งabsentabsentsuperscriptsubscript๐‘–๐‘—๐‘subscript๐ธ๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘—superscript๐นโ€ฒ๐‘งsubscript๐‘ž๐‘–๐‘—subscript~๐ป๐‘–๐‘—๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐œ‡๐‘–๐‘ฅsubscript๐œ‡๐‘–๐‘ฅsubscript๐œ‡๐‘–๐‘ฅsubscript๐’ฎ๐‘–๐‘–๐‘˜subscript๐‘ฅsubscript๐‘ž๐‘–\begin{array}[]{c}\displaystyle{\partial_{t}{U}^{\hbox{\tiny{2dSpin}}}(z)-k% \partial_{x}{V}(z)+[{U}^{\hbox{\tiny{2dSpin}}}(z),{V}(z)]=}\\ \displaystyle{=\sum\limits_{i\neq j}^{N}E_{ij}\Big{(}b\mathcal{S}_{ij}F^{% \prime}(z,q_{ij})+{\tilde{H}}_{ij}f(z,q_{ij})\Big{)}\Big{(}\mu_{i}(x)-\mu_{i}(% x)\Big{)}\,,}\\ \mu_{i}(x)=\mathcal{S}_{ii}-k\partial_{x}q_{i}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_V ( italic_z ) + [ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) , italic_V ( italic_z ) ] = end_CELL end_ROW start_ROW start_CELL = โˆ‘ start_POSTSUBSCRIPT italic_i โ‰  italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_E start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_b caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) ( italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) - italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) ) , end_CELL end_ROW start_ROW start_CELL italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (4.5)

and provides the following set of equations:

โˆ‚tpiโˆ’kโขโˆ‚xd~i+bโขโˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โ€ฒโข(qiโขm)โˆ’โˆ‘mโ‰ iN(๐’ฎiโขmโขH~mโขiโˆ’H~iโขmโข๐’ฎmโขi)โขโ„˜โข(qiโขm)=0,subscript๐‘กsubscript๐‘๐‘–๐‘˜subscript๐‘ฅsubscript~๐‘‘๐‘–๐‘superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘šsuperscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript~๐ป๐‘š๐‘–subscript~๐ป๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘š0\begin{array}[]{c}\displaystyle{\partial_{t}p_{i}-k\partial_{x}{\tilde{d}}_{i}% +b\sum\limits_{m\neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}\wp^{\prime}(q_{im}% )-\sum\limits_{m\neq i}^{N}\Big{(}\mathcal{S}_{im}{\tilde{H}}_{mi}-{\tilde{H}}% _{im}\mathcal{S}_{mi}\Big{)}\wp(q_{im})=0\,,}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_b โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT - over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) = 0 , end_CELL end_ROW end_ARRAY (4.6)
โˆ‚t๐’ฎiโขjโˆ’kโขโˆ‚xH~iโขj+2โขbโข(๐’ฎiโขiโˆ’๐’ฎjโขj)โข๐’ฎiโขjโขโ„˜โข(qiโขj)+(piโˆ’pj)โขH~iโขjโˆ’(d~iโˆ’d~j)โข๐’ฎiโขj++โˆ‘mโ‰ i,jN((๐’ฎiโขmโขH~mโขjโˆ’H~iโขmโข๐’ฎmโขj)โขฮทโข(qi,qm,qj)โˆ’bโข๐’ฎiโขmโข๐’ฎmโขjโข(โ„˜โข(qiโขm)โˆ’โ„˜โข(qmโขj)))=0forโขiโ‰ j,subscript๐‘กsubscript๐’ฎ๐‘–๐‘—๐‘˜subscript๐‘ฅsubscript~๐ป๐‘–๐‘—2๐‘subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘—๐‘—subscript๐’ฎ๐‘–๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘—subscript๐‘๐‘–subscript๐‘๐‘—subscript~๐ป๐‘–๐‘—limit-fromsubscript~๐‘‘๐‘–subscript~๐‘‘๐‘—subscript๐’ฎ๐‘–๐‘—formulae-sequencesuperscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript~๐ป๐‘š๐‘—subscript~๐ป๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘šWeierstrass-psubscript๐‘ž๐‘š๐‘—0for๐‘–๐‘—\begin{array}[]{c}\displaystyle{\partial_{t}\mathcal{S}_{ij}-k\partial_{x}{% \tilde{H}}_{ij}+2b(\mathcal{S}_{ii}-\mathcal{S}_{jj})\mathcal{S}_{ij}\wp(q_{ij% })+(p_{i}-p_{j}){\tilde{H}}_{ij}-({\tilde{d}}_{i}-{\tilde{d}}_{j})\mathcal{S}_% {ij}+}\\ \displaystyle{+\sum\limits_{m\neq i,j}^{N}\Big{(}(\mathcal{S}_{im}{\tilde{H}}_% {mj}-{\tilde{H}}_{im}\mathcal{S}_{mj})\eta(q_{i},q_{m},q_{j})-b\mathcal{S}_{im% }\mathcal{S}_{mj}(\wp(q_{im})-\wp(q_{mj}))\Big{)}=0\quad\hbox{for}\ i\neq j\,,% }\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + 2 italic_b ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - ( over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + end_CELL end_ROW start_ROW start_CELL + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT - over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - italic_b caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) ) ) = 0 for italic_i โ‰  italic_j , end_CELL end_ROW end_ARRAY (4.7)
โˆ‚t๐’ฎiโขiโˆ’kโขโˆ‚xH~iโขi=0,subscript๐‘กsubscript๐’ฎ๐‘–๐‘–๐‘˜subscript๐‘ฅsubscript~๐ป๐‘–๐‘–0\begin{array}[]{c}\displaystyle{\partial_{t}\mathcal{S}_{ii}-k\partial_{x}{% \tilde{H}}_{ii}=0\,,}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = 0 , end_CELL end_ROW end_ARRAY (4.8)
โˆ‚tqi=H~iโขi+gโข(t,x)subscript๐‘กsubscript๐‘ž๐‘–subscript~๐ป๐‘–๐‘–๐‘”๐‘ก๐‘ฅ\begin{array}[]{c}\displaystyle{\partial_{t}q_{i}={\tilde{H}}_{ii}+g(t,x)}\end% {array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + italic_g ( italic_t , italic_x ) end_CELL end_ROW end_ARRAY (4.9)

and

bโขkโขโˆ‚x๐’ฎ=[๐’ฎ,H~โˆ’bโขP],P=diagโข(p1,โ€ฆ,pN).formulae-sequence๐‘๐‘˜subscript๐‘ฅ๐’ฎ๐’ฎ~๐ป๐‘๐‘ƒ๐‘ƒdiagsubscript๐‘1โ€ฆsubscript๐‘๐‘\begin{array}[]{c}\displaystyle{bk\partial_{x}\mathcal{S}=[\mathcal{S},{\tilde% {H}}-bP]\,,\quad P={\rm diag}(p_{1},...,p_{N})\,.}\end{array}start_ARRAY start_ROW start_CELL italic_b italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S = [ caligraphic_S , over~ start_ARG italic_H end_ARG - italic_b italic_P ] , italic_P = roman_diag ( italic_p start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , โ€ฆ , italic_p start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ) . end_CELL end_ROW end_ARRAY (4.10)

In (4.7) the function ฮทโข(qi,qm,qj)๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\eta(q_{i},q_{m},q_{j})italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) from (3.8) is used.

Proof. The proof is by straightforward calculation using identities from the Appendix. The diagonal part of (4.5) provides (4.8) as coefficient behind E1โข(z)subscript๐ธ1๐‘งE_{1}(z)italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ), the diagonal part of (4.10) as coefficient behind โ„˜โข(z)Weierstrass-p๐‘ง\wp(z)โ„˜ ( italic_z ) and the equation (4.6) as a constant (in z๐‘งzitalic_z) term. Expressions in the off-diagonal part of (4.5) are expanded in linear combination of the following (linearly independent) functions: ฯ•โข(z,qiโขj)italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—\phi(z,q_{ij})italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ), fโข(z,qiโขj)๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—f(z,q_{ij})italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ), E1โข(z)โขฯ•โข(z,qiโขj)subscript๐ธ1๐‘งitalic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—E_{1}(z)\phi(z,q_{ij})italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ), E1โข(z)โขfโข(z,qiโขj)subscript๐ธ1๐‘ง๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—E_{1}(z)f(z,q_{ij})italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) and (E12โข(z)โˆ’โ„˜โข(z))โขฯ•โข(z,qiโขj)superscriptsubscript๐ธ12๐‘งWeierstrass-p๐‘งitalic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—(E_{1}^{2}(z)-\wp(z))\phi(z,q_{ij})( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) - โ„˜ ( italic_z ) ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ). The coefficients behind (E12โข(z)โˆ’โ„˜โข(z))โขฯ•โข(z,qiโขj)superscriptsubscript๐ธ12๐‘งWeierstrass-p๐‘งitalic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—(E_{1}^{2}(z)-\wp(z))\phi(z,q_{ij})( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) - โ„˜ ( italic_z ) ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) and E1โข(z)โขfโข(z,qiโขj)subscript๐ธ1๐‘ง๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—E_{1}(z)f(z,q_{ij})italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) are identities due to definition of ฮผiโข(x)subscript๐œ‡๐‘–๐‘ฅ\mu_{i}(x)italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ), the coefficient behind E1โข(z)โขฯ•โข(z,qiโขj)subscript๐ธ1๐‘งitalic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—E_{1}(z)\phi(z,q_{ij})italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) provides the off-diagonal part of (4.10), the coefficient behind ฯ•โข(z,qiโขj)italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—\phi(z,q_{ij})italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) provides (4.7) and the coefficient behind fโข(z,qiโขj)๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—f(z,q_{ij})italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) provides (4.9). โ– โ– \scriptstyle\blacksquareโ– 

In what follows we fix gโข(t,x)=0๐‘”๐‘ก๐‘ฅ0g(t,x)=0italic_g ( italic_t , italic_x ) = 0 in (4.9). Notice also that (4.8) with H~iโขi=qห™isubscript~๐ป๐‘–๐‘–subscriptห™๐‘ž๐‘–{\tilde{H}}_{ii}={\dot{q}}_{i}over~ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT is in agreement with the constraint ฮผiโข(x)=csubscript๐œ‡๐‘–๐‘ฅ๐‘\mu_{i}(x)=citalic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = italic_c since (4.8) is rewritten as โˆ‚tฮผiโข(x)=0subscript๐‘กsubscript๐œ‡๐‘–๐‘ฅ0\partial_{t}\mu_{i}(x)=0โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = 0.

The equations of motion (4.6), (4.7), (4.9) are written in terms of unknown matrix H~~๐ป{\tilde{H}}over~ start_ARG italic_H end_ARG, which should be defined from (4.10). Consider HโˆˆMatโข(N,โ„‚)๐ปMat๐‘โ„‚H\in{\rm Mat}(N,\mathbb{C})italic_H โˆˆ roman_Mat ( italic_N , blackboard_C ) satisfying

bโขkโขโˆ‚x๐’ฎ=[๐’ฎ,H].๐‘๐‘˜subscript๐‘ฅ๐’ฎ๐’ฎ๐ป\begin{array}[]{c}\displaystyle{bk\partial_{x}\mathcal{S}=[\mathcal{S},H]\,.}% \end{array}start_ARRAY start_ROW start_CELL italic_b italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S = [ caligraphic_S , italic_H ] . end_CELL end_ROW end_ARRAY (4.11)

Then H~=H+bโขP~๐ป๐ป๐‘๐‘ƒ{\tilde{H}}=H+bPover~ start_ARG italic_H end_ARG = italic_H + italic_b italic_P. Equation (4.11) appears also in the studies of higher rank Landau-Lifshitz and 1+1 Gaudin type models (see e.g. [3, 15]). A possible way to solve it is to impose condition

๐’ฎ2=c0โข๐’ฎ,superscript๐’ฎ2subscript๐‘0๐’ฎ\begin{array}[]{c}\displaystyle{{\mathcal{S}}^{2}=c_{0}{\mathcal{S}}\,,}\end{array}start_ARRAY start_ROW start_CELL caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S , end_CELL end_ROW end_ARRAY (4.12)

where c0โˆˆโ„‚subscript๐‘0โ„‚c_{0}\in\mathbb{C}italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT โˆˆ blackboard_C is some constant. Relation (4.12) means that the eigenvalues of matrix S๐‘†Sitalic_S are equal to either 00 or c0subscript๐‘0c_{0}italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT. From (4.12) it follows that ๐’ฎxโข๐’ฎ+๐’ฎโข๐’ฎx=c0โข๐’ฎxsubscript๐’ฎ๐‘ฅ๐’ฎ๐’ฎsubscript๐’ฎ๐‘ฅsubscript๐‘0subscript๐’ฎ๐‘ฅ\mathcal{S}_{x}\mathcal{S}+\mathcal{S}\mathcal{S}_{x}=c_{0}\mathcal{S}_{x}caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S + caligraphic_S caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT (where ๐’ฎx=โˆ‚x๐’ฎsubscript๐’ฎ๐‘ฅsubscript๐‘ฅ๐’ฎ\mathcal{S}_{x}=\partial_{x}\mathcal{S}caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT = โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S). Then it is easy to verify that (4.11) has solution

H=bโขkc02โข[๐’ฎ,๐’ฎx],๐ป๐‘๐‘˜superscriptsubscript๐‘02๐’ฎsubscript๐’ฎ๐‘ฅ\begin{array}[]{c}\displaystyle{H=\frac{bk}{c_{0}^{2}}\,[\mathcal{S},\mathcal{% S}_{x}]\,,}\end{array}start_ARRAY start_ROW start_CELL italic_H = divide start_ARG italic_b italic_k end_ARG start_ARG italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_ARG [ caligraphic_S , caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ] , end_CELL end_ROW end_ARRAY (4.13)

that is

H~=bโขP+bโขkc02โข[๐’ฎ,๐’ฎx].~๐ป๐‘๐‘ƒ๐‘๐‘˜superscriptsubscript๐‘02๐’ฎsubscript๐’ฎ๐‘ฅ\begin{array}[]{c}\displaystyle{{\tilde{H}}=bP+\frac{bk}{c_{0}^{2}}\,[\mathcal% {S},\mathcal{S}_{x}]\,.}\end{array}start_ARRAY start_ROW start_CELL over~ start_ARG italic_H end_ARG = italic_b italic_P + divide start_ARG italic_b italic_k end_ARG start_ARG italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_ARG [ caligraphic_S , caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ] . end_CELL end_ROW end_ARRAY (4.14)

Its diagonal part provides via (4.9) (with gโข(t,x)=0๐‘”๐‘ก๐‘ฅ0g(t,x)=0italic_g ( italic_t , italic_x ) = 0)

qห™i=bโขpi+bโขkc02โข[๐’ฎ,๐’ฎx]iโขisubscriptห™๐‘ž๐‘–๐‘subscript๐‘๐‘–๐‘๐‘˜superscriptsubscript๐‘02subscript๐’ฎsubscript๐’ฎ๐‘ฅ๐‘–๐‘–\begin{array}[]{c}\displaystyle{{\dot{q}}_{i}=bp_{i}+\frac{bk}{c_{0}^{2}}\,[% \mathcal{S},\mathcal{S}_{x}]_{ii}}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_b italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + divide start_ARG italic_b italic_k end_ARG start_ARG italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_ARG [ caligraphic_S , caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ] start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (4.15)

and the off-diagonal part of H~~๐ป\tilde{H}over~ start_ARG italic_H end_ARG enters equations (4.6)-(4.7).

The diagonal elements d~isubscript~๐‘‘๐‘–{\tilde{d}}_{i}over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT remain undefined. At the same time we of course need to reproduce the finite-dimensional equations (1.3)-(1.4) when all the fields are independent of x๐‘ฅxitalic_x (equivalently, when k=0๐‘˜0k=0italic_k = 0). When k=0๐‘˜0k=0italic_k = 0 we have H=0๐ป0H=0italic_H = 0, H~=bโขP~๐ป๐‘๐‘ƒ{\tilde{H}}=bPover~ start_ARG italic_H end_ARG = italic_b italic_P and Uspinโข(z)=Lspinโข(z)superscript๐‘ˆspin๐‘งsuperscript๐ฟspin๐‘งU^{\hbox{\tiny{spin}}}(z)=L^{\hbox{\tiny{spin}}}(z)italic_U start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) = italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ). If d~iโข(k=0)=0subscript~๐‘‘๐‘–๐‘˜00{\tilde{d}}_{i}(k=0)=0over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_k = 0 ) = 0 then Vโข(z)=E1โข(z)โขLspinโข(z)+Mspinโข(z)๐‘‰๐‘งsubscript๐ธ1๐‘งsuperscript๐ฟspin๐‘งsuperscript๐‘€spin๐‘งV(z)=E_{1}(z)L^{\hbox{\tiny{spin}}}(z)+M^{\hbox{\tiny{spin}}}(z)italic_V ( italic_z ) = italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) + italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) with b~=b~๐‘๐‘\tilde{b}=bover~ start_ARG italic_b end_ARG = italic_b. Obviously, redefinition Mspinโข(z)โ†’E1โข(z)โขLspinโข(z)+Mspinโข(z)โ†’superscript๐‘€spin๐‘งsubscript๐ธ1๐‘งsuperscript๐ฟspin๐‘งsuperscript๐‘€spin๐‘งM^{\hbox{\tiny{spin}}}(z)\rightarrow E_{1}(z)L^{\hbox{\tiny{spin}}}(z)+M^{% \hbox{\tiny{spin}}}(z)italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) โ†’ italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) + italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) does not effect the Lax equation (1.5). In this way we come to (1.3)-(1.4).

5 Specifying the ansatz for V๐‘‰Vitalic_V-matrix

There is a problem with the above given description. It is not in agreement with the spinless case (3.4)-(3.10). As will be explained below, the equation for qห™isubscriptห™๐‘ž๐‘–{\dot{q}}_{i}overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT (4.15) can not be transformed to (3.9) through the reduction to the spinless model. Another disagreement is that for Vโข(z)๐‘‰๐‘งV(z)italic_V ( italic_z ) (4.4)

Vโข(z)|k=0=E1โข(z)โขLspinโข(z)+Mspinโข(z)|b~=b,evaluated-at๐‘‰๐‘ง๐‘˜0subscript๐ธ1๐‘งsuperscript๐ฟspin๐‘งevaluated-atsuperscript๐‘€spin๐‘ง~๐‘๐‘\begin{array}[]{c}\displaystyle{V(z)|_{k=0}=E_{1}(z)L^{\hbox{\tiny{spin}}}(z)+% M^{\hbox{\tiny{spin}}}(z)|_{\tilde{b}=b}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_V ( italic_z ) | start_POSTSUBSCRIPT italic_k = 0 end_POSTSUBSCRIPT = italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_L start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) + italic_M start_POSTSUPERSCRIPT spin end_POSTSUPERSCRIPT ( italic_z ) | start_POSTSUBSCRIPT over~ start_ARG italic_b end_ARG = italic_b end_POSTSUBSCRIPT , end_CELL end_ROW end_ARRAY (5.1)

while

V2dCMโข(z)|k=0=(LCMโข(z))2โˆ’2Nโข(โˆ‘m=0Npm)โขLCMโข(z)+MCMโข(z)|b~=โˆ’2โขcevaluated-atsuperscript๐‘‰2dCM๐‘ง๐‘˜0superscriptsuperscript๐ฟCM๐‘ง22๐‘superscriptsubscript๐‘š0๐‘subscript๐‘๐‘šsuperscript๐ฟCM๐‘งevaluated-atsuperscript๐‘€CM๐‘ง~๐‘2๐‘\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dCM}}}(z)|_{k=0}=\Big{(}L^{% \hbox{\tiny{CM}}}(z)\Big{)}^{2}-\frac{2}{N}\Big{(}\sum\limits_{m=0}^{N}p_{m}% \Big{)}L^{\hbox{\tiny{CM}}}(z)+M^{\hbox{\tiny{CM}}}(z)|_{\tilde{b}=-2c}}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT ( italic_z ) | start_POSTSUBSCRIPT italic_k = 0 end_POSTSUBSCRIPT = ( italic_L start_POSTSUPERSCRIPT CM end_POSTSUPERSCRIPT ( italic_z ) ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT - divide start_ARG 2 end_ARG start_ARG italic_N end_ARG ( โˆ‘ start_POSTSUBSCRIPT italic_m = 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT ) italic_L start_POSTSUPERSCRIPT CM end_POSTSUPERSCRIPT ( italic_z ) + italic_M start_POSTSUPERSCRIPT CM end_POSTSUPERSCRIPT ( italic_z ) | start_POSTSUBSCRIPT over~ start_ARG italic_b end_ARG = - 2 italic_c end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (5.2)

up to constant matrices proportional to the identity matrix 1Nsubscript1๐‘1_{N}1 start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT.

5.1 Another role of condition ๐’ฎ2=c0โข๐’ฎsuperscript๐’ฎ2subscript๐‘0๐’ฎ{\mathcal{S}}^{2}=c_{0}{\mathcal{S}}caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S

Besides solving (4.11) we are going to use the condition (4.12) to modify the definition of V๐‘‰Vitalic_V-matrix.

Proposition 1

Suppose the condition (4.12) holds true. Then the matrix V0โข(z)โˆˆMatโข(N,โ„‚)superscript๐‘‰0๐‘งMat๐‘โ„‚V^{0}(z)\in{\rm Mat}(N,\mathbb{C})italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT ( italic_z ) โˆˆ roman_Mat ( italic_N , blackboard_C )

Viโขj0โข(z)=โˆ’[(U2dSpinโข(z))2]iโขj++ฮดiโขjโข(๐’ฎiโขi)2โข(E12โข(z)โˆ’โ„˜โข(z))+(1โˆ’ฮดiโขj)โข(๐’ฎiโขi+๐’ฎjโขj)โข๐’ฎiโขjโขfโข(z,qiโขj)subscriptsuperscript๐‘‰0๐‘–๐‘—๐‘งlimit-fromsubscriptdelimited-[]superscriptsuperscript๐‘ˆ2dSpin๐‘ง2๐‘–๐‘—missing-subexpressionsubscript๐›ฟ๐‘–๐‘—superscriptsubscript๐’ฎ๐‘–๐‘–2superscriptsubscript๐ธ12๐‘งWeierstrass-p๐‘ง1subscript๐›ฟ๐‘–๐‘—subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘—๐‘—subscript๐’ฎ๐‘–๐‘—๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{V^{0}_{ij}(z)=-\Big{[}\Big{(}U^{\hbox{\tiny{2% dSpin}}}(z)\Big{)}^{2}\Big{]}_{ij}+}\\ \\ \displaystyle{+\delta_{ij}(\mathcal{S}_{ii})^{2}\Big{(}E_{1}^{2}(z)-\wp(z)\Big% {)}+(1-\delta_{ij})(\mathcal{S}_{ii}+\mathcal{S}_{jj})\mathcal{S}_{ij}f(z,q_{% ij})}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = - [ ( italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ] start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) - โ„˜ ( italic_z ) ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (5.3)

has the form of the matrix Vโข(z)๐‘‰๐‘งV(z)italic_V ( italic_z ) (4.4) with

b=โˆ’c0.๐‘subscript๐‘0\begin{array}[]{c}\displaystyle{b=-c_{0}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_b = - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (5.4)

Proof. The following statement is verified directly:

[(U2dSpinโข(z))2]iโขj==ฮดiโขjโข(pi2โˆ’โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โข(qiโขm)+O~iโขiโขE1โข(z)+(๐’ฎiโขi)2โข(E12โข(z)โˆ’โ„˜โข(z))+(๐’ฎ2)iโขiโขโ„˜โข(z))++(1โˆ’ฮดiโขj)โข((๐’ฎ2)iโขjโขFโข(z,qiโขj)+O~iโขjโขฯ•โข(z,qiโขj)+(๐’ฎiโขi+๐’ฎjโขj)โข๐’ฎiโขjโขfโข(z,qiโขj)),subscriptdelimited-[]superscriptsuperscript๐‘ˆ2dSpin๐‘ง2๐‘–๐‘—absentmissing-subexpressionabsentlimit-fromsubscript๐›ฟ๐‘–๐‘—superscriptsubscript๐‘๐‘–2superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘šsubscript~๐‘‚๐‘–๐‘–subscript๐ธ1๐‘งsuperscriptsubscript๐’ฎ๐‘–๐‘–2superscriptsubscript๐ธ12๐‘งWeierstrass-p๐‘งsubscriptsuperscript๐’ฎ2๐‘–๐‘–Weierstrass-p๐‘งmissing-subexpression1subscript๐›ฟ๐‘–๐‘—subscriptsuperscript๐’ฎ2๐‘–๐‘—๐น๐‘งsubscript๐‘ž๐‘–๐‘—subscript~๐‘‚๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘—๐‘—subscript๐’ฎ๐‘–๐‘—๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{\Big{[}\Big{(}U^{\hbox{\tiny{2dSpin}}}(z)\Big{% )}^{2}\Big{]}_{ij}=}\\ \\ \displaystyle{=\delta_{ij}\Big{(}p_{i}^{2}-\sum\limits_{m\neq i}^{N}\mathcal{S% }_{im}\mathcal{S}_{mi}\wp(q_{im})+{\tilde{O}}_{ii}E_{1}(z)+(\mathcal{S}_{ii})^% {2}\Big{(}E_{1}^{2}(z)-\wp(z)\Big{)}+(\mathcal{S}^{2})_{ii}\wp(z)\Big{)}+}\\ \\ \displaystyle{+(1-\delta_{ij})\Big{(}(\mathcal{S}^{2})_{ij}F(z,q_{ij})+{\tilde% {O}}_{ij}\phi(z,q_{ij})+(\mathcal{S}_{ii}+\mathcal{S}_{jj})\mathcal{S}_{ij}f(z% ,q_{ij})\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL [ ( italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ] start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) + over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) - โ„˜ ( italic_z ) ) + ( caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT โ„˜ ( italic_z ) ) + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( ( caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (5.5)

where the matrix O~โˆˆMatโข(N,โ„‚)~๐‘‚Mat๐‘โ„‚{\tilde{O}}\in{\rm Mat}(N,\mathbb{C})over~ start_ARG italic_O end_ARG โˆˆ roman_Mat ( italic_N , blackboard_C ) is as follows:

O~iโขj=ฮดiโขjโข2โขpiโข๐’ฎiโขi+(1โˆ’ฮดiโขj)โข((pi+pj)โข๐’ฎiโขj+โˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโขฮทโข(qi,qm,qj)),subscript~๐‘‚๐‘–๐‘—subscript๐›ฟ๐‘–๐‘—2subscript๐‘๐‘–subscript๐’ฎ๐‘–๐‘–1subscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–subscript๐‘๐‘—subscript๐’ฎ๐‘–๐‘—superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{{\tilde{O}}_{ij}=\delta_{ij}2p_{i}\mathcal{S}_% {ii}+(1-\delta_{ij})\Big{(}(p_{i}+p_{j})\mathcal{S}_{ij}+\sum\limits_{m\neq i,% j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}\eta(q_{i},q_{m},q_{j})\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT 2 italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (5.6)

where ฮทโข(qi,qm,qj)๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\eta(q_{i},q_{m},q_{j})italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) is the function (3.8). Plugging ๐’ฎ2=c0โข๐’ฎsuperscript๐’ฎ2subscript๐‘0๐’ฎ{\mathcal{S}}^{2}=c_{0}{\mathcal{S}}caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S we get:

Viโขj0โข(z)=โˆ’ฮดiโขjโข(pi2โˆ’โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โข(qiโขm)+O~iโขiโขE1โข(z)+c0โข๐’ฎiโขiโขโ„˜โข(z))โˆ’โˆ’(1โˆ’ฮดiโขj)โข(c0โข๐’ฎiโขjโขFโข(z,qiโขj)+O~iโขjโขฯ•โข(z,qiโขj)).subscriptsuperscript๐‘‰0๐‘–๐‘—๐‘งlimit-fromsubscript๐›ฟ๐‘–๐‘—superscriptsubscript๐‘๐‘–2superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘šsubscript~๐‘‚๐‘–๐‘–subscript๐ธ1๐‘งsubscript๐‘0subscript๐’ฎ๐‘–๐‘–Weierstrass-p๐‘งmissing-subexpression1subscript๐›ฟ๐‘–๐‘—subscript๐‘0subscript๐’ฎ๐‘–๐‘—๐น๐‘งsubscript๐‘ž๐‘–๐‘—subscript~๐‘‚๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{V^{0}_{ij}(z)=-\delta_{ij}\Big{(}p_{i}^{2}-% \sum\limits_{m\neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}\wp(q_{im})+{\tilde{O% }}_{ii}E_{1}(z)+c_{0}\mathcal{S}_{ii}\wp(z)\Big{)}-}\\ \\ \displaystyle{-(1-\delta_{ij})\Big{(}c_{0}\mathcal{S}_{ij}F(z,q_{ij})+{\tilde{% O}}_{ij}\phi(z,q_{ij})\Big{)}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) + over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT โ„˜ ( italic_z ) ) - end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL - ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) . end_CELL end_ROW end_ARRAY (5.7)

This finishes the proof. โ– โ– \scriptstyle\blacksquareโ– 

The defined above matrix O~~๐‘‚{\tilde{O}}over~ start_ARG italic_O end_ARG (5.6) has an important property.

Proposition 2

Suppose the condition (4.12) holds true. Then

[๐’ฎ,O~โˆ’c0โขP]=0,๐’ฎ~๐‘‚subscript๐‘0๐‘ƒ0\begin{array}[]{c}\displaystyle{[\mathcal{S},{\tilde{O}}-c_{0}P]=0\,,}\end{array}start_ARRAY start_ROW start_CELL [ caligraphic_S , over~ start_ARG italic_O end_ARG - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_P ] = 0 , end_CELL end_ROW end_ARRAY (5.8)

where P=diagโข(p1,โ€ฆ,pN)๐‘ƒdiagsubscript๐‘1โ€ฆsubscript๐‘๐‘P={\rm diag}(p_{1},...,p_{N})italic_P = roman_diag ( italic_p start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , โ€ฆ , italic_p start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ).

Proof. The proof is based on skew-symmetry of the function ฮทโข(qi,qm,qj)๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\eta(q_{i},q_{m},q_{j})italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) (3.8) with respect to permutation of any pair of arguments. The latter follows from (A.5)-(A.6). Also, it is convenient to use a simple relation

ฮทโข(qm,ql,qj)โˆ’ฮทโข(qi,qm,ql)=ฮทโข(qi,ql,qj)โˆ’ฮทโข(qi,qm,qj),๐œ‚subscript๐‘ž๐‘šsubscript๐‘ž๐‘™subscript๐‘ž๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘™๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘™subscript๐‘ž๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{\eta(q_{m},q_{l},q_{j})-\eta(q_{i},q_{m},q_{l}% )=\eta(q_{i},q_{l},q_{j})-\eta(q_{i},q_{m},q_{j})\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_l end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_l end_POSTSUBSCRIPT ) = italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_l end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) - italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) , end_CELL end_ROW end_ARRAY (5.9)

which follows from the definition (3.8). โ– โ– \scriptstyle\blacksquareโ– 

In fact, the relation (5.8) can be proved indirectly. We have two different representations (5.3) and (5.7) for the matrix V0โข(z)superscript๐‘‰0๐‘งV^{0}(z)italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT ( italic_z ). One can use both representations to compute the commutator [U2dSpin,V0โข(z)]superscript๐‘ˆ2dSpinsuperscript๐‘‰0๐‘ง[U^{\hbox{\tiny{2dSpin}}},V^{0}(z)][ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT , italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT ( italic_z ) ]. For computation of the commutator in representation (5.7) one should use identities (A.7)-(A.11), while for the representation (5.3) formula (A.15) should be used. By comparing both calculations one gets two different representations for certain expressions. In this way (5.8) follows as well as two more nontrivial relations.

Proposition 3

Suppose the condition (4.12) is valid. Then the following relations hold true:

c0โขโˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โ€ฒโข(qiโขm)โˆ’โˆ‘mโ‰ iN(๐’ฎiโขmโขO~mโขiโˆ’O~iโขmโข๐’ฎmโขi)โขโ„˜โข(qiโขm)=โˆ’โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโข(๐’ฎiโขi+๐’ฎmโขm)โขโ„˜โ€ฒโข(qiโขm)subscript๐‘0superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘šsuperscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript~๐‘‚๐‘š๐‘–subscript~๐‘‚๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘šsuperscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘š๐‘šsuperscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘š\begin{array}[]{c}\displaystyle{c_{0}\sum\limits_{m\neq i}^{N}\mathcal{S}_{im}% \mathcal{S}_{mi}\wp^{\prime}(q_{im})-\sum\limits_{m\neq i}^{N}\Big{(}\mathcal{% S}_{im}{\tilde{O}}_{mi}-{\tilde{O}}_{im}\mathcal{S}_{mi}\Big{)}\wp(q_{im})=-% \sum\limits_{m\neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}(\mathcal{S}_{ii}+% \mathcal{S}_{mm})\wp^{\prime}(q_{im})}\end{array}start_ARRAY start_ROW start_CELL italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT - over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) = - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (5.10)

for any i=1,โ€ฆ,N๐‘–1โ€ฆ๐‘i=1,...,Nitalic_i = 1 , โ€ฆ , italic_N and

(piโˆ’pj)โขO~iโขjโˆ’(diโˆ’dj)โข๐’ฎiโขj+โˆ‘mโ‰ i,jN(๐’ฎiโขmโขO~mโขjโˆ’O~iโขmโข๐’ฎmโขj)โขฮทโข(qi,qm,qj)โˆ’c0โขโˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโข(โ„˜โข(qiโขm)โˆ’โ„˜โข(qmโขj))=(๐’ฎiโขiโˆ’๐’ฎjโขj)โขโ„˜โข(qiโขj)โขโˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขj++โˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโข((๐’ฎiโขi+๐’ฎmโขm)โขโ„˜โข(qiโขm)โˆ’(๐’ฎjโขj+๐’ฎmโขm)โขโ„˜โข(qjโขm))forโขiโ‰ j.subscript๐‘๐‘–subscript๐‘๐‘—subscript~๐‘‚๐‘–๐‘—subscript๐‘‘๐‘–subscript๐‘‘๐‘—subscript๐’ฎ๐‘–๐‘—superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript~๐‘‚๐‘š๐‘—subscript~๐‘‚๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—subscript๐‘0superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘šWeierstrass-psubscript๐‘ž๐‘š๐‘—limit-fromsubscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘—๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘—superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘š๐‘šWeierstrass-psubscript๐‘ž๐‘–๐‘šsubscript๐’ฎ๐‘—๐‘—subscript๐’ฎ๐‘š๐‘šWeierstrass-psubscript๐‘ž๐‘—๐‘šfor๐‘–๐‘—\begin{array}[]{c}\displaystyle{(p_{i}-p_{j}){\tilde{O}}_{ij}-(d_{i}-d_{j})% \mathcal{S}_{ij}+\sum\limits_{m\neq i,j}^{N}\Big{(}\mathcal{S}_{im}{\tilde{O}}% _{mj}-{\tilde{O}}_{im}\mathcal{S}_{mj}\Big{)}\eta(q_{i},q_{m},q_{j})}\\ \displaystyle{-c_{0}\sum\limits_{m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj% }(\wp(q_{im})-\wp(q_{mj}))=(\mathcal{S}_{ii}-\mathcal{S}_{jj})\wp(q_{ij})\sum% \limits_{m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}+}\\ \displaystyle{+\sum\limits_{m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}\Big% {(}(\mathcal{S}_{ii}+\mathcal{S}_{mm})\wp(q_{im})-(\mathcal{S}_{jj}+\mathcal{S% }_{mm})\wp(q_{jm})\Big{)}\quad\hbox{for}\ i\neq j\,.}\end{array}start_ARRAY start_ROW start_CELL ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - ( italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_d start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT - over~ start_ARG italic_O end_ARG start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) end_CELL end_ROW start_ROW start_CELL - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) ) = ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT + end_CELL end_ROW start_ROW start_CELL + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - ( caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j italic_m end_POSTSUBSCRIPT ) ) for italic_i โ‰  italic_j . end_CELL end_ROW end_ARRAY (5.11)

We will use these relations below to transform equations of motion.

5.2 The final answer for V๐‘‰Vitalic_V-matrix and equations of motion

Introduce the matrix V2dSpinโˆˆMatโข(N,โ„‚)superscript๐‘‰2dSpinMat๐‘โ„‚V^{\hbox{\tiny{2dSpin}}}\in{\rm Mat}(N,\mathbb{C})italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT โˆˆ roman_Mat ( italic_N , blackboard_C ) with elements

Viโขj2dSpinโข(z)=Viโขj0โข(z)โˆ’2โขฮบโขUiโขj2dSpinโข(z)+ฮดiโขjโขHiโขiโขE1โข(z)+(1โˆ’ฮดiโขj)โขHiโขjโขฯ•โข(z,qiโขj)subscriptsuperscript๐‘‰2dSpin๐‘–๐‘—๐‘งsubscriptsuperscript๐‘‰0๐‘–๐‘—๐‘ง2๐œ…subscriptsuperscript๐‘ˆ2dSpin๐‘–๐‘—๐‘งsubscript๐›ฟ๐‘–๐‘—subscript๐ป๐‘–๐‘–subscript๐ธ1๐‘ง1subscript๐›ฟ๐‘–๐‘—subscript๐ป๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dSpin}}}_{ij}(z)=V^{0}_{ij}(z)% -2\kappa U^{\hbox{\tiny{2dSpin}}}_{ij}(z)+\delta_{ij}H_{ii}E_{1}(z)+(1-\delta_% {ij})H_{ij}\phi(z,q_{ij})}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) = italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) - 2 italic_ฮบ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_z ) + italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_H start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_H start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (5.12)

or, equivalently,

V2dSpinโข(z)=โˆ’ฮดiโขjโข(di+OiโขiโขE1โข(z)+c0โข๐’ฎiโขiโขโ„˜โข(z))โˆ’(1โˆ’ฮดiโขj)โข(c0โข๐’ฎiโขjโขFโข(z,qiโขj)+Oiโขjโขฯ•โข(z,qiโขj))++ฮดiโขjโขHiโขiโขE1โข(z)+(1โˆ’ฮดiโขj)โขHiโขjโขฯ•โข(z,qiโขj),superscript๐‘‰2dSpin๐‘งsubscript๐›ฟ๐‘–๐‘—subscript๐‘‘๐‘–subscript๐‘‚๐‘–๐‘–subscript๐ธ1๐‘งsubscript๐‘0subscript๐’ฎ๐‘–๐‘–Weierstrass-p๐‘งlimit-from1subscript๐›ฟ๐‘–๐‘—subscript๐‘0subscript๐’ฎ๐‘–๐‘—๐น๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐‘‚๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—missing-subexpressionsubscript๐›ฟ๐‘–๐‘—subscript๐ป๐‘–๐‘–subscript๐ธ1๐‘ง1subscript๐›ฟ๐‘–๐‘—subscript๐ป๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dSpin}}}(z)=-\delta_{ij}\Big{(% }d_{i}+{O}_{ii}E_{1}(z)+c_{0}\mathcal{S}_{ii}\wp(z)\Big{)}-(1-\delta_{ij})\Big% {(}c_{0}\mathcal{S}_{ij}F(z,q_{ij})+{O}_{ij}\phi(z,q_{ij})\Big{)}+}\\ \\ \displaystyle{+\delta_{ij}H_{ii}E_{1}(z)+(1-\delta_{ij})H_{ij}\phi(z,q_{ij})\,% ,}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) = - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_O start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT โ„˜ ( italic_z ) ) - ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + italic_O start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_H start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_H start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) , end_CELL end_ROW end_ARRAY (5.13)

where

di=pi2+2โขฮบโขpiโˆ’โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โข(qiโขm),i=1,โ€ฆ,Nformulae-sequencesubscript๐‘‘๐‘–superscriptsubscript๐‘๐‘–22๐œ…subscript๐‘๐‘–superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘š๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{d_{i}=p_{i}^{2}+2\kappa p_{i}-\sum\limits_{m% \neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}\wp(q_{im})\,,\quad i=1,...,N}\end{array}start_ARRAY start_ROW start_CELL italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + 2 italic_ฮบ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) , italic_i = 1 , โ€ฆ , italic_N end_CELL end_ROW end_ARRAY (5.14)

and the matrix O๐‘‚Oitalic_O is defined as

Oiโขj=ฮดiโขjโข2โข(pi+ฮบ)โข๐’ฎiโขi+(1โˆ’ฮดiโขj)โข((pi+pj+2โขฮบ)โข๐’ฎiโขj+โˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโขฮทโข(qi,qm,qj)).subscript๐‘‚๐‘–๐‘—subscript๐›ฟ๐‘–๐‘—2subscript๐‘๐‘–๐œ…subscript๐’ฎ๐‘–๐‘–1subscript๐›ฟ๐‘–๐‘—subscript๐‘๐‘–subscript๐‘๐‘—2๐œ…subscript๐’ฎ๐‘–๐‘—superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—\begin{array}[]{c}\displaystyle{{O}_{ij}=\delta_{ij}2(p_{i}+\kappa)\mathcal{S}% _{ii}+(1-\delta_{ij})\Big{(}(p_{i}+p_{j}+2\kappa)\mathcal{S}_{ij}+\sum\limits_% {m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}\eta(q_{i},q_{m},q_{j})\Big{)}% \,.}\end{array}start_ARRAY start_ROW start_CELL italic_O start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT 2 ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT + 2 italic_ฮบ ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) ) . end_CELL end_ROW end_ARRAY (5.15)

The variable ฮบ๐œ…\kappaitalic_ฮบ is free, it will be fixed below.

Notice that similarly to O~~๐‘‚\tilde{O}over~ start_ARG italic_O end_ARG the matrix O๐‘‚Oitalic_O satisfies the property

[๐’ฎ,Oโˆ’c0โขP]=0๐’ฎ๐‘‚subscript๐‘0๐‘ƒ0\begin{array}[]{c}\displaystyle{[\mathcal{S},{O}-c_{0}P]=0}\end{array}start_ARRAY start_ROW start_CELL [ caligraphic_S , italic_O - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_P ] = 0 end_CELL end_ROW end_ARRAY (5.16)

due to O๐‘‚Oitalic_O is obtained from O~~๐‘‚\tilde{O}over~ start_ARG italic_O end_ARG by the shift piโ†’pi+ฮบโ†’subscript๐‘๐‘–subscript๐‘๐‘–๐œ…p_{i}\rightarrow p_{i}+\kappaitalic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT โ†’ italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ: O=O~+2โขฮบโข๐’ฎ๐‘‚~๐‘‚2๐œ…๐’ฎO={\tilde{O}}+2\kappa\mathcal{S}italic_O = over~ start_ARG italic_O end_ARG + 2 italic_ฮบ caligraphic_S, and O~~๐‘‚\tilde{O}over~ start_ARG italic_O end_ARG satisfies (5.8).

Let us formulate the main statement.

Theorem 2

Suppose the condition (4.12) holds true. The U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair (4.1) and (5.13) satisfies the Zakharov-Shabat equation with the unwanted term

โˆ‚tU2dSpinโข(z)โˆ’kโขโˆ‚xV2dSpinโข(z)+[U2dSpinโข(z),V2dSpinโข(z)]==โˆ‘iโ‰ jNEiโขjโข(โˆ’c0โข๐’ฎiโขjโขFโ€ฒโข(z,qiโขj)+Hห˜iโขjโขfโข(z,qiโขj))โข(ฮผiโข(x)โˆ’ฮผiโข(x)),subscript๐‘กsuperscript๐‘ˆ2dSpin๐‘ง๐‘˜subscript๐‘ฅsuperscript๐‘‰2dSpin๐‘งsuperscript๐‘ˆ2dSpin๐‘งsuperscript๐‘‰2dSpin๐‘งabsentmissing-subexpressionabsentsuperscriptsubscript๐‘–๐‘—๐‘subscript๐ธ๐‘–๐‘—subscript๐‘0subscript๐’ฎ๐‘–๐‘—superscript๐นโ€ฒ๐‘งsubscript๐‘ž๐‘–๐‘—subscriptห˜๐ป๐‘–๐‘—๐‘“๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐œ‡๐‘–๐‘ฅsubscript๐œ‡๐‘–๐‘ฅ\begin{array}[]{c}\displaystyle{\partial_{t}{U}^{\hbox{\tiny{2dSpin}}}(z)-k% \partial_{x}{V}^{\hbox{\tiny{2dSpin}}}(z)+[{U}^{\hbox{\tiny{2dSpin}}}(z),{V}^{% \hbox{\tiny{2dSpin}}}(z)]=}\\ \\ \displaystyle{=\sum\limits_{i\neq j}^{N}E_{ij}\Big{(}-c_{0}\mathcal{S}_{ij}F^{% \prime}(z,q_{ij})+{\breve{H}}_{ij}f(z,q_{ij})\Big{)}\Big{(}\mu_{i}(x)-\mu_{i}(% x)\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) + [ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) , italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) ] = end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL = โˆ‘ start_POSTSUBSCRIPT italic_i โ‰  italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_E start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_f ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) ( italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) - italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) ) , end_CELL end_ROW end_ARRAY (5.17)

where

ฮผiโข(x)=๐’ฎiโขiโˆ’kโขโˆ‚xqi,Hห˜=Hโˆ’OโˆˆMatโข(N,โ„‚)subscript๐œ‡๐‘–๐‘ฅsubscript๐’ฎ๐‘–๐‘–๐‘˜subscript๐‘ฅsubscript๐‘ž๐‘–missing-subexpressionห˜๐ป๐ป๐‘‚Mat๐‘โ„‚\begin{array}[]{c}\displaystyle{\mu_{i}(x)=\mathcal{S}_{ii}-k\partial_{x}q_{i}% \,,}\\ \\ {\breve{H}}=H-O\in{\rm Mat}(N,\mathbb{C})\end{array}start_ARRAY start_ROW start_CELL italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL overห˜ start_ARG italic_H end_ARG = italic_H - italic_O โˆˆ roman_Mat ( italic_N , blackboard_C ) end_CELL end_ROW end_ARRAY (5.18)

and provides the following set of equations:

โˆ‚tpi+kโขโˆ‚xdiโˆ’c0โขโˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโขโ„˜โ€ฒโข(qiโขm)โˆ’โˆ‘mโ‰ iN(๐’ฎiโขmโขHห˜mโขiโˆ’Hห˜iโขmโข๐’ฎmโขi)โขโ„˜โข(qiโขm)=0,subscript๐‘กsubscript๐‘๐‘–๐‘˜subscript๐‘ฅsubscript๐‘‘๐‘–subscript๐‘0superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘šsuperscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscriptห˜๐ป๐‘š๐‘–subscriptห˜๐ป๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘š0\begin{array}[]{c}\displaystyle{\partial_{t}p_{i}+k\partial_{x}{d}_{i}-c_{0}% \sum\limits_{m\neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}\wp^{\prime}(q_{im})-% \sum\limits_{m\neq i}^{N}\Big{(}\mathcal{S}_{im}{\breve{H}}_{mi}-{\breve{H}}_{% im}\mathcal{S}_{mi}\Big{)}\wp(q_{im})=0\,,}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT - overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) = 0 , end_CELL end_ROW end_ARRAY (5.19)
โˆ‚t๐’ฎiโขjโˆ’kโขโˆ‚xHห˜iโขjโˆ’2โขc0โข(๐’ฎiโขiโˆ’๐’ฎjโขj)โข๐’ฎiโขjโขโ„˜โข(qiโขj)+(piโˆ’pj)โขHห˜iโขj+(diโˆ’dj)โข๐’ฎiโขj++โˆ‘mโ‰ i,jN((๐’ฎiโขmโขHห˜mโขjโˆ’Hห˜iโขmโข๐’ฎmโขj)โขฮทโข(qi,qm,qj)+c0โข๐’ฎiโขmโข๐’ฎmโขjโข(โ„˜โข(qiโขm)โˆ’โ„˜โข(qmโขj)))=0forโขiโ‰ j,subscript๐‘กsubscript๐’ฎ๐‘–๐‘—๐‘˜subscript๐‘ฅsubscriptห˜๐ป๐‘–๐‘—2subscript๐‘0subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘—๐‘—subscript๐’ฎ๐‘–๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘—subscript๐‘๐‘–subscript๐‘๐‘—subscriptห˜๐ป๐‘–๐‘—limit-fromsubscript๐‘‘๐‘–subscript๐‘‘๐‘—subscript๐’ฎ๐‘–๐‘—formulae-sequencesuperscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐’ฎ๐‘–๐‘šsubscriptห˜๐ป๐‘š๐‘—subscriptห˜๐ป๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—๐œ‚subscript๐‘ž๐‘–subscript๐‘ž๐‘šsubscript๐‘ž๐‘—subscript๐‘0subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘šWeierstrass-psubscript๐‘ž๐‘š๐‘—0for๐‘–๐‘—\begin{array}[]{c}\displaystyle{\partial_{t}\mathcal{S}_{ij}-k\partial_{x}{% \breve{H}}_{ij}-2c_{0}(\mathcal{S}_{ii}-\mathcal{S}_{jj})\mathcal{S}_{ij}\wp(q% _{ij})+(p_{i}-p_{j}){\breve{H}}_{ij}+({d}_{i}-{d}_{j})\mathcal{S}_{ij}+}\\ \displaystyle{+\sum\limits_{m\neq i,j}^{N}\Big{(}(\mathcal{S}_{im}{\breve{H}}_% {mj}-{\breve{H}}_{im}\mathcal{S}_{mj})\eta(q_{i},q_{m},q_{j})+c_{0}\mathcal{S}% _{im}\mathcal{S}_{mj}(\wp(q_{im})-\wp(q_{mj}))\Big{)}=0\quad\hbox{for}\ i\neq j% \,,}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - 2 italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + ( italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_d start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + end_CELL end_ROW start_ROW start_CELL + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT - overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) + italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) ) ) = 0 for italic_i โ‰  italic_j , end_CELL end_ROW end_ARRAY (5.20)
โˆ‚t๐’ฎiโขiโˆ’kโขโˆ‚xHห˜iโขi=0,subscript๐‘กsubscript๐’ฎ๐‘–๐‘–๐‘˜subscript๐‘ฅsubscriptห˜๐ป๐‘–๐‘–0\begin{array}[]{c}\displaystyle{\partial_{t}\mathcal{S}_{ii}-k\partial_{x}{% \breve{H}}_{ii}=0\,,}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = 0 , end_CELL end_ROW end_ARRAY (5.21)
โˆ‚tqi=Hห˜iโขisubscript๐‘กsubscript๐‘ž๐‘–subscriptห˜๐ป๐‘–๐‘–\begin{array}[]{c}\displaystyle{\partial_{t}q_{i}={\breve{H}}_{ii}}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (5.22)

and

โˆ’c0โขkโขโˆ‚x๐’ฎ=[๐’ฎ,Hห˜+c0โขP],P=diagโข(p1,โ€ฆ,pN).formulae-sequencesubscript๐‘0๐‘˜subscript๐‘ฅ๐’ฎ๐’ฎห˜๐ปsubscript๐‘0๐‘ƒ๐‘ƒdiagsubscript๐‘1โ€ฆsubscript๐‘๐‘\begin{array}[]{c}\displaystyle{-c_{0}k\partial_{x}\mathcal{S}=[\mathcal{S},{% \breve{H}}+c_{0}P]\,,\quad P={\rm diag}(p_{1},...,p_{N})\,.}\end{array}start_ARRAY start_ROW start_CELL - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S = [ caligraphic_S , overห˜ start_ARG italic_H end_ARG + italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_P ] , italic_P = roman_diag ( italic_p start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , โ€ฆ , italic_p start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ) . end_CELL end_ROW end_ARRAY (5.23)

The matrix O๐‘‚Oitalic_O entering the definition of Hห˜ห˜๐ป{\breve{H}}overห˜ start_ARG italic_H end_ARG (5.18) is defined in (5.15), and disubscript๐‘‘๐‘–d_{i}italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT in (5.14). Also, the matrix H๐ปHitalic_H from (5.18) is the solution (4.13) of (4.11) with b=โˆ’c0๐‘subscript๐‘0b=-c_{0}italic_b = - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT (5.4), that is

H=โˆ’kc0โข[๐’ฎ,๐’ฎx].๐ป๐‘˜subscript๐‘0๐’ฎsubscript๐’ฎ๐‘ฅ\begin{array}[]{c}\displaystyle{H=-\frac{k}{c_{0}}\,[\mathcal{S},\mathcal{S}_{% x}]\,.}\end{array}start_ARRAY start_ROW start_CELL italic_H = - divide start_ARG italic_k end_ARG start_ARG italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT end_ARG [ caligraphic_S , caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ] . end_CELL end_ROW end_ARRAY (5.24)

The equations (5.19)-(5.20) can be rewritten in the following way:

โˆ‚tpi+kโขโˆ‚xdi+โˆ‘mโ‰ iN๐’ฎiโขmโข๐’ฎmโขiโข(๐’ฎiโขi+๐’ฎmโขm)โขโ„˜โ€ฒโข(qiโขm)โˆ’โˆ‘mโ‰ iN(๐’ฎiโขmโขHmโขiโˆ’Hiโขmโข๐’ฎmโขi)โขโ„˜โข(qiโขm)=0subscript๐‘กsubscript๐‘๐‘–๐‘˜subscript๐‘ฅsubscript๐‘‘๐‘–superscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–subscript๐’ฎ๐‘–๐‘–subscript๐’ฎ๐‘š๐‘šsuperscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘šsuperscriptsubscript๐‘š๐‘–๐‘subscript๐’ฎ๐‘–๐‘šsubscript๐ป๐‘š๐‘–subscript๐ป๐‘–๐‘šsubscript๐’ฎ๐‘š๐‘–Weierstrass-psubscript๐‘ž๐‘–๐‘š0\begin{array}[]{c}\displaystyle{\partial_{t}p_{i}+k\partial_{x}{d}_{i}+\sum% \limits_{m\neq i}^{N}\mathcal{S}_{im}\mathcal{S}_{mi}(\mathcal{S}_{ii}+% \mathcal{S}_{mm})\wp^{\prime}(q_{im})-\sum\limits_{m\neq i}^{N}\Big{(}\mathcal% {S}_{im}{H}_{mi}-{H}_{im}\mathcal{S}_{mi}\Big{)}\wp(q_{im})=0}\end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT italic_H start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT - italic_H start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) = 0 end_CELL end_ROW end_ARRAY (5.25)

and

โˆ‚t๐’ฎiโขjโˆ’kโขโˆ‚xHห˜iโขjโˆ’2โขc0โข(๐’ฎiโขiโˆ’๐’ฎjโขj)โข๐’ฎiโขjโขโ„˜โข(qiโขj)++(piโˆ’pj)Hiโขj+โˆ‘mโ‰ i,jN((๐’ฎiโขmHmโขjโˆ’Hiโขm๐’ฎmโขj)ฮท(qi,qm,qj)++โˆ‘mโ‰ i,jN๐’ฎiโขmโข๐’ฎmโขjโข((๐’ฎiโขi+๐’ฎmโขm)โขโ„˜โข(qiโขm)โˆ’(๐’ฎjโขj+๐’ฎmโขm)โขโ„˜โข(qjโขm)+(๐’ฎiโขiโˆ’๐’ฎjโขj)โขโ„˜โข(qiโขj))=0\begin{array}[]{c}\displaystyle{\partial_{t}\mathcal{S}_{ij}-k\partial_{x}{% \breve{H}}_{ij}-2c_{0}(\mathcal{S}_{ii}-\mathcal{S}_{jj})\mathcal{S}_{ij}\wp(q% _{ij})+}\\ \\ \displaystyle{+(p_{i}-p_{j})H_{ij}+\sum\limits_{m\neq i,j}^{N}\Big{(}(\mathcal% {S}_{im}{H}_{mj}-{H}_{im}\mathcal{S}_{mj})\eta(q_{i},q_{m},q_{j})+}\\ \displaystyle{+\sum\limits_{m\neq i,j}^{N}\mathcal{S}_{im}\mathcal{S}_{mj}\Big% {(}(\mathcal{S}_{ii}+\mathcal{S}_{mm})\wp(q_{im})-(\mathcal{S}_{jj}+\mathcal{S% }_{mm})\wp(q_{jm})+(\mathcal{S}_{ii}-\mathcal{S}_{jj})\wp(q_{ij})\Big{)}=0}% \end{array}start_ARRAY start_ROW start_CELL โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT overห˜ start_ARG italic_H end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - 2 italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - italic_p start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) italic_H start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT italic_H start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT - italic_H start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ) italic_ฮท ( italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT , italic_q start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ) + end_CELL end_ROW start_ROW start_CELL + โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - ( caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT + caligraphic_S start_POSTSUBSCRIPT italic_m italic_m end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j italic_m end_POSTSUBSCRIPT ) + ( caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - caligraphic_S start_POSTSUBSCRIPT italic_j italic_j end_POSTSUBSCRIPT ) โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) = 0 end_CELL end_ROW end_ARRAY (5.26)

for iโ‰ j๐‘–๐‘—i\neq jitalic_i โ‰  italic_j and i,j=1,โ€ฆ,Nformulae-sequence๐‘–๐‘—1โ€ฆ๐‘i,j=1,...,Nitalic_i , italic_j = 1 , โ€ฆ , italic_N.

Proof. Let us combine similar terms in V2dSpinโข(z)superscript๐‘‰2dSpin๐‘ง{V}^{\hbox{\tiny{2dSpin}}}(z)italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) (5.13):

V2dSpinโข(z)=ฮดiโขjโข(โˆ’di+(Hiโขiโˆ’Oiโขi)โขE1โข(z)โˆ’c0โข๐’ฎiโขiโขโ„˜โข(z))++(1โˆ’ฮดiโขj)โข(โˆ’c0โข๐’ฎiโขjโขFโข(z,qiโขj)+(Hiโขjโˆ’Oiโขj)โขฯ•โข(z,qiโขj)).superscript๐‘‰2dSpin๐‘งlimit-fromsubscript๐›ฟ๐‘–๐‘—subscript๐‘‘๐‘–subscript๐ป๐‘–๐‘–subscript๐‘‚๐‘–๐‘–subscript๐ธ1๐‘งsubscript๐‘0subscript๐’ฎ๐‘–๐‘–Weierstrass-p๐‘งmissing-subexpression1subscript๐›ฟ๐‘–๐‘—subscript๐‘0subscript๐’ฎ๐‘–๐‘—๐น๐‘งsubscript๐‘ž๐‘–๐‘—subscript๐ป๐‘–๐‘—subscript๐‘‚๐‘–๐‘—italic-ฯ•๐‘งsubscript๐‘ž๐‘–๐‘—\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dSpin}}}(z)=\delta_{ij}\Big{(}% -d_{i}+({H}_{ii}-O_{ii})E_{1}(z)-c_{0}\mathcal{S}_{ii}\wp(z)\Big{)}+}\\ \\ \displaystyle{+(1-\delta_{ij})\Big{(}-c_{0}\mathcal{S}_{ij}F(z,q_{ij})+(H_{ij}% -{O}_{ij})\phi(z,q_{ij})\Big{)}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ( - italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + ( italic_H start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - italic_O start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT ) italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT โ„˜ ( italic_z ) ) + end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL + ( 1 - italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ( - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT italic_F ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) + ( italic_H start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT - italic_O start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z , italic_q start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT ) ) . end_CELL end_ROW end_ARRAY (5.27)

Then we apply the Theorem 1, where bโ†’โˆ’c0โ†’๐‘subscript๐‘0b\rightarrow-c_{0}italic_b โ†’ - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT, H~โ†’Hห˜=Hโˆ’Oโ†’~๐ปห˜๐ป๐ป๐‘‚{\tilde{H}}\rightarrow{\breve{H}}=H-Oover~ start_ARG italic_H end_ARG โ†’ overห˜ start_ARG italic_H end_ARG = italic_H - italic_O and d~iโ†’โˆ’diโ†’subscript~๐‘‘๐‘–subscript๐‘‘๐‘–{\tilde{d}}_{i}\rightarrow-d_{i}over~ start_ARG italic_d end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT โ†’ - italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT. This yields (5.19)-(5.23). Notice that due to Proposition 2 and (5.16) the relation (5.23) transforms as:

โˆ’c0โขkโขโˆ‚x๐’ฎ=[๐’ฎ,Hห˜+c0โขP]=[๐’ฎ,H+c0โขPโˆ’O]=[๐’ฎ,H],subscript๐‘0๐‘˜subscript๐‘ฅ๐’ฎ๐’ฎห˜๐ปsubscript๐‘0๐‘ƒ๐’ฎ๐ปsubscript๐‘0๐‘ƒ๐‘‚๐’ฎ๐ป\begin{array}[]{c}\displaystyle{-c_{0}k\partial_{x}\mathcal{S}=[\mathcal{S},{% \breve{H}}+c_{0}P]=[\mathcal{S},H+c_{0}P-O]=[\mathcal{S},H]\,,}\end{array}start_ARRAY start_ROW start_CELL - italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S = [ caligraphic_S , overห˜ start_ARG italic_H end_ARG + italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_P ] = [ caligraphic_S , italic_H + italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_P - italic_O ] = [ caligraphic_S , italic_H ] , end_CELL end_ROW end_ARRAY (5.28)

that is H๐ปHitalic_H is indeed given by (5.24). Next, consider equation (5.22)666It is, in fact, includes an arbitrary function gโข(t,x)๐‘”๐‘ก๐‘ฅg(t,x)italic_g ( italic_t , italic_x ) similarly to (4.9) which we again fix to be zero.. Due to (5.18), (5.15) and (5.24) from (5.22) we have

qห™i=โˆ’2โข(pi+ฮบ)โข๐’ฎiโขiโˆ’kc0โข[๐’ฎ,๐’ฎx]iโขi,i=1,โ€ฆ,N.formulae-sequencesubscriptห™๐‘ž๐‘–2subscript๐‘๐‘–๐œ…subscript๐’ฎ๐‘–๐‘–๐‘˜subscript๐‘0subscript๐’ฎsubscript๐’ฎ๐‘ฅ๐‘–๐‘–๐‘–1โ€ฆ๐‘\begin{array}[]{c}\displaystyle{{\dot{q}}_{i}=-2(p_{i}+\kappa)\mathcal{S}_{ii}% -\frac{k}{c_{0}}\,[\mathcal{S},\mathcal{S}_{x}]_{ii}\,,\quad i=1,...,N\,.}\end% {array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = - 2 ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ ) caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT - divide start_ARG italic_k end_ARG start_ARG italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT end_ARG [ caligraphic_S , caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ] start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT , italic_i = 1 , โ€ฆ , italic_N . end_CELL end_ROW end_ARRAY (5.29)

Notice that this expression for qห™isubscriptห™๐‘ž๐‘–{\dot{q}}_{i}overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT is different from (4.15) and is similar to (3.9).

Finally, the equations (5.19)-(5.20) are rewritten as (5.25)-(5.26) by using (5.10) and (5.11) from the Proposition 3. A remark is that the relations (5.10) and (5.11) are written for the matrix O~~๐‘‚\tilde{O}over~ start_ARG italic_O end_ARG, and they are valid for the matrix O๐‘‚Oitalic_O as well since transition O~โ†’Oโ†’~๐‘‚๐‘‚\tilde{O}\rightarrow Oover~ start_ARG italic_O end_ARG โ†’ italic_O in these formulae corresponds to computation in two ways of [U2dSpin,V0โข(z)โˆ’2โขฮบโขU2dSpin]=[U2dSpin,V0โข(z)]superscript๐‘ˆ2dSpinsuperscript๐‘‰0๐‘ง2๐œ…superscript๐‘ˆ2dSpinsuperscript๐‘ˆ2dSpinsuperscript๐‘‰0๐‘ง[U^{\hbox{\tiny{2dSpin}}},V^{0}(z)-2\kappa U^{\hbox{\tiny{2dSpin}}}]=[U^{\hbox% {\tiny{2dSpin}}},V^{0}(z)][ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT , italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT ( italic_z ) - 2 italic_ฮบ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ] = [ italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT , italic_V start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT ( italic_z ) ] as was explained after Proposition 2. โ– โ– \scriptstyle\blacksquareโ– 

Let us now fix the parameter ฮบ๐œ…\kappaitalic_ฮบ. We choose it in such a way that โˆ‘i=1Nqห™i=0superscriptsubscript๐‘–1๐‘subscriptห™๐‘ž๐‘–0\sum\limits_{i=1}^{N}{\dot{q}}_{i}=0โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT overห™ start_ARG italic_q end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = 0. Summing up equations (5.29) we obtain

ฮบ=โˆ’1trโข(๐’ฎ)โขโˆ‘i=1N๐’ฎiโขiโขpi.๐œ…1tr๐’ฎsuperscriptsubscript๐‘–1๐‘subscript๐’ฎ๐‘–๐‘–subscript๐‘๐‘–\begin{array}[]{c}\displaystyle{\kappa=-\frac{1}{{\rm tr}(\mathcal{S})}\sum% \limits_{i=1}^{N}\mathcal{S}_{ii}p_{i}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_ฮบ = - divide start_ARG 1 end_ARG start_ARG roman_tr ( caligraphic_S ) end_ARG โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (5.30)

Notice also that with this choice the sum โˆ‘i=1Nqisuperscriptsubscript๐‘–1๐‘subscript๐‘ž๐‘–\sum\limits_{i=1}^{N}{q}_{i}โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT is a function of x๐‘ฅxitalic_x. It depends on initial conditions. Let us fix it be equal to zero:

โˆ‘i=1Nqi=0.superscriptsubscript๐‘–1๐‘subscript๐‘ž๐‘–0\begin{array}[]{c}\displaystyle{\sum\limits_{i=1}^{N}{q}_{i}=0\,.}\end{array}start_ARRAY start_ROW start_CELL โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = 0 . end_CELL end_ROW end_ARRAY (5.31)

At the same time the sum of momenta is not zero. This is similar to the spinless case [1].

It follows from (5.21) and (5.22) that โˆ‚tฮผiโข(x)=0subscript๐‘กsubscript๐œ‡๐‘–๐‘ฅ0\partial_{t}\mu_{i}(x)=0โˆ‚ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = 0. In principle, one may fix ฮผiโข(x)subscript๐œ‡๐‘–๐‘ฅ\mu_{i}(x)italic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) to be a function of x๐‘ฅxitalic_x, but we choose the level of the moment map to be independent of x๐‘ฅxitalic_x: ฮผiโข(x)=csubscript๐œ‡๐‘–๐‘ฅ๐‘\mu_{i}(x)=citalic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = italic_c.

Consistency with the finite-dimensional model.

The form (5.25)-(5.26) of equations of motion is useful for the limit to finite-dimensional case. In this limit all the fields become x๐‘ฅxitalic_x-independent variables. That is all derivatives with respect to x๐‘ฅxitalic_x vanishes, or equivalently, one should put k=0๐‘˜0k=0italic_k = 0. The constraints ฮผiโข(x)=csubscript๐œ‡๐‘–๐‘ฅ๐‘\mu_{i}(x)=citalic_ฮผ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_x ) = italic_c turn into Siโขi=csubscript๐‘†๐‘–๐‘–๐‘S_{ii}=citalic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = italic_c (the latter follows from (5.21)). Then the equations (5.25)-(5.26) on-shell the conditions Siโขi=csubscript๐‘†๐‘–๐‘–๐‘S_{ii}=citalic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = italic_c are as follows:

pห™i=โˆ’2โขcโขโˆ‘mโ‰ iNSiโขmโขSmโขiโขโ„˜โ€ฒโข(qiโขm)subscriptห™๐‘๐‘–2๐‘superscriptsubscript๐‘š๐‘–๐‘subscript๐‘†๐‘–๐‘šsubscript๐‘†๐‘š๐‘–superscriptWeierstrass-pโ€ฒsubscript๐‘ž๐‘–๐‘š\begin{array}[]{c}\displaystyle{{\dot{p}}_{i}=-2c\sum\limits_{m\neq i}^{N}S_{% im}S_{mi}\wp^{\prime}(q_{im})}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_p end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = - 2 italic_c โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT italic_S start_POSTSUBSCRIPT italic_m italic_i end_POSTSUBSCRIPT โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (5.32)

and

Sห™iโขj=โˆ’2โขcโขโˆ‘mโ‰ i,jNSiโขmโขSmโขjโข(โ„˜โข(qiโขm)โˆ’โ„˜โข(qjโขm)),subscriptห™๐‘†๐‘–๐‘—2๐‘superscriptsubscript๐‘š๐‘–๐‘—๐‘subscript๐‘†๐‘–๐‘šsubscript๐‘†๐‘š๐‘—Weierstrass-psubscript๐‘ž๐‘–๐‘šWeierstrass-psubscript๐‘ž๐‘—๐‘š\begin{array}[]{c}\displaystyle{{\dot{S}}_{ij}=-2c\sum\limits_{m\neq i,j}^{N}S% _{im}S_{mj}(\wp(q_{im})-\wp(q_{jm}))\,,}\end{array}start_ARRAY start_ROW start_CELL overห™ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = - 2 italic_c โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i , italic_j end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_S start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT italic_S start_POSTSUBSCRIPT italic_m italic_j end_POSTSUBSCRIPT ( โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - โ„˜ ( italic_q start_POSTSUBSCRIPT italic_j italic_m end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (5.33)

which are equations (1.4) with b~=โˆ’2โขc~๐‘2๐‘\tilde{b}=-2cover~ start_ARG italic_b end_ARG = - 2 italic_c.

6 Reduction to the spinless field Calogero-Moser model

Here we show that the obtained equations of motion for the spin field Calogero-Moser model are in agreement with the equations for the spinless model (1.7)-(1.8) from [1]. For this purpose we perform a reduction to the spinless model. Since the Hamiltonian is unknown in the field case, our strategy is to use equations of motion similarly to the calculation (2.12)-(2.16).

Suppose

๐’ฎiโขj=ฮพiโขฯˆj.subscript๐’ฎ๐‘–๐‘—subscript๐œ‰๐‘–subscript๐œ“๐‘—\begin{array}[]{c}\displaystyle{\mathcal{S}_{ij}=\xi_{i}\psi_{j}\,.}\end{array}start_ARRAY start_ROW start_CELL caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (6.1)

The reduction includes the constraints ๐’ฎiโขi=ฮฑi2subscript๐’ฎ๐‘–๐‘–superscriptsubscript๐›ผ๐‘–2\mathcal{S}_{ii}=\alpha_{i}^{2}caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT = italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT (4.3) together with some gauge fixation, which we choose to be ฮพi=ฯˆisubscript๐œ‰๐‘–subscript๐œ“๐‘–\xi_{i}=\psi_{i}italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_ฯˆ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, that is

onshell:ฮพi=ฯˆi=(kqi,x+c)1/2=ฮฑi,i=1,โ€ฆ,N.\begin{array}[]{c}\displaystyle{{\rm on\ shell}:\ \xi_{i}=\psi_{i}=(kq_{i,x}+c% )^{1/2}=\alpha_{i}\,,\quad i=1,...,N\,.}\end{array}start_ARRAY start_ROW start_CELL roman_on roman_shell : italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = italic_ฯˆ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = ( italic_k italic_q start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT + italic_c ) start_POSTSUPERSCRIPT 1 / 2 end_POSTSUPERSCRIPT = italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_i = 1 , โ€ฆ , italic_N . end_CELL end_ROW end_ARRAY (6.2)

Then we perform the gauge transformation

U2dSpinโข(z)โ†’gโˆ’1โขU2dSpinโข(z)โขgโˆ’kโขgโˆ’1โขโˆ‚xg,โ†’superscript๐‘ˆ2dSpin๐‘งsuperscript๐‘”1superscript๐‘ˆ2dSpin๐‘ง๐‘”๐‘˜superscript๐‘”1subscript๐‘ฅ๐‘”\begin{array}[]{c}\displaystyle{U^{\hbox{\tiny{2dSpin}}}(z)\rightarrow g^{-1}U% ^{\hbox{\tiny{2dSpin}}}(z)g-kg^{-1}{\partial_{x}g}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) โ†’ italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) italic_g - italic_k italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_g , end_CELL end_ROW end_ARRAY (6.3)
V2dSpinโข(z)โ†’gโˆ’1โขV2dSpinโข(z)โขgโˆ’gโˆ’1โขgห™โ†’superscript๐‘‰2dSpin๐‘งsuperscript๐‘”1superscript๐‘‰2dSpin๐‘ง๐‘”superscript๐‘”1ห™๐‘”\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dSpin}}}(z)\rightarrow g^{-1}V% ^{\hbox{\tiny{2dSpin}}}(z)g-g^{-1}{\dot{g}}}\end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) โ†’ italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) italic_g - italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT overห™ start_ARG italic_g end_ARG end_CELL end_ROW end_ARRAY (6.4)

with g=diagโข(ฮพ1,โ€ฆ,ฮพN)๐‘”diagsubscript๐œ‰1โ€ฆsubscript๐œ‰๐‘g={\rm diag}(\xi_{1},...,\xi_{N})italic_g = roman_diag ( italic_ฮพ start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , โ€ฆ , italic_ฮพ start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ). As a result of this gauge transformation, the matrix U๐‘ˆUitalic_U depends on ๐’ฎiโขisubscript๐’ฎ๐‘–๐‘–\mathcal{S}_{ii}caligraphic_S start_POSTSUBSCRIPT italic_i italic_i end_POSTSUBSCRIPT (which is a function of qisubscript๐‘ž๐‘–q_{i}italic_q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT due to (5.21)-(5.22)) and โˆ‚xฮพi/ฮพisubscript๐‘ฅsubscript๐œ‰๐‘–subscript๐œ‰๐‘–\partial_{x}\xi_{i}/\xi_{i}โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT / italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT. Therefore, the spin variables are gauged away similarly to the finite-dimensional case, and the values of โˆ‚xฮพi/ฮพisubscript๐‘ฅsubscript๐œ‰๐‘–subscript๐œ‰๐‘–\partial_{x}\xi_{i}/\xi_{i}โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT / italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT depend on the gauge fixation. The U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair for the spinless model appear as

U2dCMโข(z)=(gโˆ’1โขU2dSpinโข(z)โขgโˆ’kโขgโˆ’1โขโˆ‚xg)|onโขshellsuperscript๐‘ˆ2dCM๐‘งevaluated-atsuperscript๐‘”1superscript๐‘ˆ2dSpin๐‘ง๐‘”๐‘˜superscript๐‘”1subscript๐‘ฅ๐‘”onshell\begin{array}[]{c}\displaystyle{U^{\hbox{\tiny{2dCM}}}(z)=\Big{(}g^{-1}U^{% \hbox{\tiny{2dSpin}}}(z)g-kg^{-1}{\partial_{x}g}\Big{)}\Big{|}_{\rm on\ shell}% }\end{array}start_ARRAY start_ROW start_CELL italic_U start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT ( italic_z ) = ( italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_U start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) italic_g - italic_k italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_g ) | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (6.5)

and

V2dCMโข(z)=(gโˆ’1โขV2dSpinโข(z)โขgโˆ’gโˆ’1โขgห™)|onโขshell.superscript๐‘‰2dCM๐‘งevaluated-atsuperscript๐‘”1superscript๐‘‰2dSpin๐‘ง๐‘”superscript๐‘”1ห™๐‘”onshell\begin{array}[]{c}\displaystyle{V^{\hbox{\tiny{2dCM}}}(z)=\Big{(}g^{-1}V^{% \hbox{\tiny{2dSpin}}}(z)g-g^{-1}{\dot{g}}\Big{)}\Big{|}_{\rm on\ shell}\,.}% \end{array}start_ARRAY start_ROW start_CELL italic_V start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT ( italic_z ) = ( italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT italic_V start_POSTSUPERSCRIPT 2dSpin end_POSTSUPERSCRIPT ( italic_z ) italic_g - italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT overห™ start_ARG italic_g end_ARG ) | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (6.6)

Since

(kโขgโˆ’1โขโˆ‚xg)iโขj=ฮดiโขjโขkโขฮพi,xฮพi+kโขฮปxsubscript๐‘˜superscript๐‘”1subscript๐‘ฅ๐‘”๐‘–๐‘—subscript๐›ฟ๐‘–๐‘—๐‘˜subscript๐œ‰๐‘–๐‘ฅsubscript๐œ‰๐‘–๐‘˜subscript๐œ†๐‘ฅ\begin{array}[]{c}\displaystyle{\Big{(}kg^{-1}{\partial_{x}g}\Big{)}_{ij}=% \delta_{ij}\frac{k\xi_{i,x}}{\xi_{i}}+k\lambda_{x}}\end{array}start_ARRAY start_ROW start_CELL ( italic_k italic_g start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT italic_g ) start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮด start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT divide start_ARG italic_k italic_ฮพ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG + italic_k italic_ฮป start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (6.7)

it is easy to see that the expression in the r.h.s. of (6.6) coincides with (3.4) up to kโขฮปx๐‘˜subscript๐œ†๐‘ฅk\lambda_{x}italic_k italic_ฮป start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT, which comes from the freedom in definition of ฮพisubscript๐œ‰๐‘–\xi_{i}italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, ฯˆjsubscript๐œ“๐‘—\psi_{j}italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT (ฮพiโ†’eฮปโขฮพiโ†’subscript๐œ‰๐‘–superscript๐‘’๐œ†subscript๐œ‰๐‘–\xi_{i}\rightarrow e^{\lambda}\xi_{i}italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT โ†’ italic_e start_POSTSUPERSCRIPT italic_ฮป end_POSTSUPERSCRIPT italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, ฯˆjโ†’eโˆ’ฮปโขฯˆjโ†’subscript๐œ“๐‘—superscript๐‘’๐œ†subscript๐œ“๐‘—\psi_{j}\rightarrow e^{-\lambda}\psi_{j}italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT โ†’ italic_e start_POSTSUPERSCRIPT - italic_ฮป end_POSTSUPERSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT). But this input can be gauged away and we put ฮป=0๐œ†0\lambda=0italic_ฮป = 0.

Due to (6.1) the matrix ๐’ฎ๐’ฎ\mathcal{S}caligraphic_S on shell the constraints (6.2) has the form

๐’ฎiโขj=ฮพiโขฯˆj|onโขshell=ฮฑiโขฮฑj.subscript๐’ฎ๐‘–๐‘—evaluated-atsubscript๐œ‰๐‘–subscript๐œ“๐‘—onshellsubscript๐›ผ๐‘–subscript๐›ผ๐‘—\begin{array}[]{c}\displaystyle{\mathcal{S}_{ij}=\xi_{i}\psi_{j}\Big{|}_{\rm on% \ shell}=\alpha_{i}\alpha_{j}\,.}\end{array}start_ARRAY start_ROW start_CELL caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฯˆ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT = italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (6.8)

Below we assume ๐’ฎiโขjsubscript๐’ฎ๐‘–๐‘—\mathcal{S}_{ij}caligraphic_S start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT to be on shell constrains. From (5.22) we conclude that

โˆ‘i=1Nฮฑi2=Nโขc,superscriptsubscript๐‘–1๐‘superscriptsubscript๐›ผ๐‘–2๐‘๐‘\begin{array}[]{c}\displaystyle{\sum\limits_{i=1}^{N}\alpha_{i}^{2}=Nc\,,}\end% {array}start_ARRAY start_ROW start_CELL โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_N italic_c , end_CELL end_ROW end_ARRAY (6.9)

and, therefore,

๐’ฎ2=Nโขcโข๐’ฎ,superscript๐’ฎ2๐‘๐‘๐’ฎ\begin{array}[]{c}\displaystyle{\mathcal{S}^{2}=Nc\mathcal{S}\,,}\end{array}start_ARRAY start_ROW start_CELL caligraphic_S start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = italic_N italic_c caligraphic_S , end_CELL end_ROW end_ARRAY (6.10)

that is

c0=Nโขc.subscript๐‘0๐‘๐‘\begin{array}[]{c}\displaystyle{c_{0}=Nc\,.}\end{array}start_ARRAY start_ROW start_CELL italic_c start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = italic_N italic_c . end_CELL end_ROW end_ARRAY (6.11)

Therefore, the function ฮบ๐œ…\kappaitalic_ฮบ turns into

ฮบ|onโขshell=โˆ’1Nโขcโขโˆ‘i=1Nฮฑi2โขpievaluated-at๐œ…onshell1๐‘๐‘superscriptsubscript๐‘–1๐‘superscriptsubscript๐›ผ๐‘–2subscript๐‘๐‘–\begin{array}[]{c}\displaystyle{\kappa\Big{|}_{\rm on\ shell}=-\frac{1}{Nc}% \sum\limits_{i=1}^{N}\alpha_{i}^{2}p_{i}}\end{array}start_ARRAY start_ROW start_CELL italic_ฮบ | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT = - divide start_ARG 1 end_ARG start_ARG italic_N italic_c end_ARG โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_CELL end_ROW end_ARRAY (6.12)

as in (3.2). Also, it follows from (6.9) that

โˆ‘i=1Nฮฑiโขฮฑi,x=0.superscriptsubscript๐‘–1๐‘subscript๐›ผ๐‘–subscript๐›ผ๐‘–๐‘ฅ0\begin{array}[]{c}\displaystyle{\sum\limits_{i=1}^{N}\alpha_{i}\alpha_{i,x}=0% \,.}\end{array}start_ARRAY start_ROW start_CELL โˆ‘ start_POSTSUBSCRIPT italic_i = 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_N end_POSTSUPERSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT = 0 . end_CELL end_ROW end_ARRAY (6.13)

Then (๐’ฎxโข๐’ฎ)iโขj=Nโขcโขฮฑฮฑi,xโขฮฑjsubscriptsubscript๐’ฎ๐‘ฅ๐’ฎ๐‘–๐‘—๐‘๐‘subscript๐›ผsubscript๐›ผ๐‘–๐‘ฅsubscript๐›ผ๐‘—(\mathcal{S}_{x}\mathcal{S})_{ij}=Nc\alpha_{\alpha_{i,x}\alpha_{j}}( caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT caligraphic_S ) start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_N italic_c italic_ฮฑ start_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT end_POSTSUBSCRIPT, (๐’ฎโข๐’ฎx)iโขj=Nโขcโขฮฑฮฑiโขฮฑj,xsubscript๐’ฎsubscript๐’ฎ๐‘ฅ๐‘–๐‘—๐‘๐‘subscript๐›ผsubscript๐›ผ๐‘–subscript๐›ผ๐‘—๐‘ฅ(\mathcal{S}\mathcal{S}_{x})_{ij}=Nc\alpha_{\alpha_{i}\alpha_{j,x}}( caligraphic_S caligraphic_S start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ) start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT = italic_N italic_c italic_ฮฑ start_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT end_POSTSUBSCRIPT and for the matrix H๐ปHitalic_H (5.24) using also (6.11) we obtain

Hiโขj|onโขshell=kโขฮฑi,xโขฮฑjโˆ’kโขฮฑiโขฮฑj,x.evaluated-atsubscript๐ป๐‘–๐‘—onshell๐‘˜subscript๐›ผ๐‘–๐‘ฅsubscript๐›ผ๐‘—๐‘˜subscript๐›ผ๐‘–subscript๐›ผ๐‘—๐‘ฅ\begin{array}[]{c}\displaystyle{H_{ij}\Big{|}_{\rm on\ shell}=k\alpha_{i,x}% \alpha_{j}-k\alpha_{i}\alpha_{j,x}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_H start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT = italic_k italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT - italic_k italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_j , italic_x end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (6.14)

To finish the derivation of U๐‘ˆUitalic_U-V๐‘‰Vitalic_V pair for the spinless model we need to show that the r.h.s. of (6.6) indeed coincides with V2dCMโข(z)superscript๐‘‰2dCM๐‘งV^{\hbox{\tiny{2dCM}}}(z)italic_V start_POSTSUPERSCRIPT 2dCM end_POSTSUPERSCRIPT ( italic_z ) (3.5)-(3.7). Most of terms are easily obtained by substitution of (6.8)-(6.14) into (5.27) except the diagonal z๐‘งzitalic_z-independent part, which is transformed under the gauge transformation as

diโ†’di+ฮพห™iฮพi|onโขshell.โ†’subscript๐‘‘๐‘–subscript๐‘‘๐‘–evaluated-atsubscriptห™๐œ‰๐‘–subscript๐œ‰๐‘–onshell\begin{array}[]{c}\displaystyle{d_{i}\rightarrow d_{i}+\frac{{\dot{\xi}}_{i}}{% \xi_{i}}\Big{|}_{\rm on\ shell}\,.}\end{array}start_ARRAY start_ROW start_CELL italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT โ†’ italic_d start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + divide start_ARG overห™ start_ARG italic_ฮพ end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT . end_CELL end_ROW end_ARRAY (6.15)

Let us compute ฮพห™iฮพi|onโขshellevaluated-atsubscriptห™๐œ‰๐‘–subscript๐œ‰๐‘–onshell\frac{{\dot{\xi}}_{i}}{\xi_{i}}\Big{|}_{\rm on\ shell}divide start_ARG overห™ start_ARG italic_ฮพ end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT. Due to (2.12) we need to compute the r.h.s. of ๐’ฎห™iโขjsubscriptห™๐’ฎ๐‘–๐‘—{\dot{\mathcal{S}}}_{ij}overห™ start_ARG caligraphic_S end_ARG start_POSTSUBSCRIPT italic_i italic_j end_POSTSUBSCRIPT on shell the constraints. Using equations of motion (5.26) it is straightforwardly verified that

ฮพห™iฮพi|onโขshell=k2โขฮฑi,xโขxฮฑiโˆ’2โข(pi+ฮบ)โขkโขฮฑi,xฮฑiโˆ’kโขโˆ‚x(pi+ฮบ)โˆ’โˆ’2โขโˆ‘mโ‰ iฮฑm4โขโ„˜โข(qiโขm)โˆ’4โขkโขโˆ‘mโ‰ iฮฑmโขฮฑm,xโขฮถโข(qiโขm).evaluated-atsubscriptห™๐œ‰๐‘–subscript๐œ‰๐‘–onshellsuperscript๐‘˜2subscript๐›ผ๐‘–๐‘ฅ๐‘ฅsubscript๐›ผ๐‘–2subscript๐‘๐‘–๐œ…๐‘˜subscript๐›ผ๐‘–๐‘ฅsubscript๐›ผ๐‘–limit-from๐‘˜subscript๐‘ฅsubscript๐‘๐‘–๐œ…missing-subexpression2subscript๐‘š๐‘–superscriptsubscript๐›ผ๐‘š4Weierstrass-psubscript๐‘ž๐‘–๐‘š4๐‘˜subscript๐‘š๐‘–subscript๐›ผ๐‘šsubscript๐›ผ๐‘š๐‘ฅ๐œsubscript๐‘ž๐‘–๐‘š\begin{array}[]{c}\displaystyle{\frac{{\dot{\xi}}_{i}}{\xi_{i}}\Big{|}_{\rm on% \ shell}=k^{2}\frac{\alpha_{i,xx}}{\alpha_{i}}-2(p_{i}+\kappa)k\frac{\alpha_{i% ,x}}{\alpha_{i}}-k\partial_{x}(p_{i}+\kappa)-}\\ \\ \displaystyle{-2\sum\limits_{m\neq i}\alpha_{m}^{4}\wp(q_{im})-4k\sum\limits_{% m\neq i}\alpha_{m}\alpha_{m,x}\zeta(q_{im})\,.}\end{array}start_ARRAY start_ROW start_CELL divide start_ARG overห™ start_ARG italic_ฮพ end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮพ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG | start_POSTSUBSCRIPT roman_on roman_shell end_POSTSUBSCRIPT = italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG - 2 ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ ) italic_k divide start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i , italic_x end_POSTSUBSCRIPT end_ARG start_ARG italic_ฮฑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG - italic_k โˆ‚ start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT ( italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ฮบ ) - end_CELL end_ROW start_ROW start_CELL end_CELL end_ROW start_ROW start_CELL - 2 โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT โ„˜ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) - 4 italic_k โˆ‘ start_POSTSUBSCRIPT italic_m โ‰  italic_i end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_m end_POSTSUBSCRIPT italic_ฮฑ start_POSTSUBSCRIPT italic_m , italic_x end_POSTSUBSCRIPT italic_ฮถ ( italic_q start_POSTSUBSCRIPT italic_i italic_m end_POSTSUBSCRIPT ) . end_CELL end_ROW end_ARRAY (6.16)

In this way the V๐‘‰Vitalic_V-matrix (3.5) is reproduced.

7 Appendix: elliptic functions

Main object is the elliptic Kronecker function [11]:

ฯ•โข(z,u)=ฯ‘โ€ฒโข(0)โขฯ‘โข(z+u)ฯ‘โข(z)โขฯ‘โข(u)=ฯ•โข(u,z),Resz=0ฯ•โข(z,u)=1,ฯ•โข(โˆ’z,โˆ’u)=โˆ’ฯ•โข(z,u),formulae-sequenceitalic-ฯ•๐‘ง๐‘ขsuperscriptitalic-ฯ‘โ€ฒ0italic-ฯ‘๐‘ง๐‘ขitalic-ฯ‘๐‘งitalic-ฯ‘๐‘ขitalic-ฯ•๐‘ข๐‘งformulae-sequencesubscriptRes๐‘ง0italic-ฯ•๐‘ง๐‘ข1italic-ฯ•๐‘ง๐‘ขitalic-ฯ•๐‘ง๐‘ข\begin{array}[]{l}\displaystyle{\phi(z,u)=\frac{\vartheta^{\prime}(0)\vartheta% (z+u)}{\vartheta(z)\vartheta(u)}=\phi(u,z)\,,\quad\mathop{\hbox{Res}}\limits% \limits_{z=0}\phi(z,u)=1\,,\quad\phi(-z,-u)=-\phi(z,u)\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z , italic_u ) = divide start_ARG italic_ฯ‘ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( 0 ) italic_ฯ‘ ( italic_z + italic_u ) end_ARG start_ARG italic_ฯ‘ ( italic_z ) italic_ฯ‘ ( italic_u ) end_ARG = italic_ฯ• ( italic_u , italic_z ) , Res start_POSTSUBSCRIPT italic_z = 0 end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_u ) = 1 , italic_ฯ• ( - italic_z , - italic_u ) = - italic_ฯ• ( italic_z , italic_u ) , end_CELL end_ROW end_ARRAY (A.1)

where ฯ‘โข(z)italic-ฯ‘๐‘ง\vartheta(z)italic_ฯ‘ ( italic_z ) is the first Jacobi theta-function. In the Riemann notations it is

ฯ‘โข(z)=ฯ‘โข(z,ฯ„)โ‰กโˆ’ฮธโข[1/21/2]โข(z|ฯ„),italic-ฯ‘๐‘งitalic-ฯ‘๐‘ง๐œ๐œƒdelimited-[]1212conditional๐‘ง๐œ\begin{array}[]{c}\displaystyle{\vartheta(z)=\vartheta(z,\tau)\equiv-\theta{% \left[\begin{array}[]{c}1/2\\ 1/2\end{array}\right]}(z|\,\tau)\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ‘ ( italic_z ) = italic_ฯ‘ ( italic_z , italic_ฯ„ ) โ‰ก - italic_ฮธ [ start_ARRAY start_ROW start_CELL 1 / 2 end_CELL end_ROW start_ROW start_CELL 1 / 2 end_CELL end_ROW end_ARRAY ] ( italic_z | italic_ฯ„ ) , end_CELL end_ROW end_ARRAY (A.2)
ฮธโข[ab]โข(z|ฯ„)=โˆ‘jโˆˆฮถexpโก(2โขฯ€โขฤฑโข(j+a)2โขฯ„2+2โขฯ€โขฤฑโข(j+a)โข(z+b)),Imโข(ฯ„)>0,formulae-sequence๐œƒdelimited-[]๐‘Ž๐‘conditional๐‘ง๐œsubscript๐‘—๐œ2๐œ‹italic-ฤฑsuperscript๐‘—๐‘Ž2๐œ22๐œ‹italic-ฤฑ๐‘—๐‘Ž๐‘ง๐‘Im๐œ0\begin{array}[]{c}\displaystyle{\theta{\left[\begin{array}[]{c}a\\ b\end{array}\right]}(z|\,\tau)=\sum_{j\in{\zeta}}\exp\left(2\pi\imath(j+a)^{2}% \frac{\tau}{2}+2\pi\imath(j+a)(z+b)\right)\,,\quad{\rm Im}(\tau)>0\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฮธ [ start_ARRAY start_ROW start_CELL italic_a end_CELL end_ROW start_ROW start_CELL italic_b end_CELL end_ROW end_ARRAY ] ( italic_z | italic_ฯ„ ) = โˆ‘ start_POSTSUBSCRIPT italic_j โˆˆ italic_ฮถ end_POSTSUBSCRIPT roman_exp ( 2 italic_ฯ€ italic_ฤฑ ( italic_j + italic_a ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT divide start_ARG italic_ฯ„ end_ARG start_ARG 2 end_ARG + 2 italic_ฯ€ italic_ฤฑ ( italic_j + italic_a ) ( italic_z + italic_b ) ) , roman_Im ( italic_ฯ„ ) > 0 , end_CELL end_ROW end_ARRAY (A.3)

where ฯ„๐œ\tauitalic_ฯ„ is the moduli of elliptic curve โ„‚/(โ„ค+ฯ„โขโ„ค)โ„‚โ„ค๐œโ„ค\mathbb{C}/(\mathbb{Z}+\tau\mathbb{Z})blackboard_C / ( blackboard_Z + italic_ฯ„ blackboard_Z ).

The partial derivative fโข(z,u)=โˆ‚uฯ†โข(z,u)๐‘“๐‘ง๐‘ขsubscript๐‘ข๐œ‘๐‘ง๐‘ขf(z,u)=\partial_{u}\varphi(z,u)italic_f ( italic_z , italic_u ) = โˆ‚ start_POSTSUBSCRIPT italic_u end_POSTSUBSCRIPT italic_ฯ† ( italic_z , italic_u ) is

fโข(z,u)=ฯ•โข(z,u)โข(E1โข(z+u)โˆ’E1โข(u)),fโข(โˆ’z,โˆ’u)=fโข(z,u)formulae-sequence๐‘“๐‘ง๐‘ขitalic-ฯ•๐‘ง๐‘ขsubscript๐ธ1๐‘ง๐‘ขsubscript๐ธ1๐‘ข๐‘“๐‘ง๐‘ข๐‘“๐‘ง๐‘ข\begin{array}[]{c}\displaystyle{\par f(z,u)=\phi(z,u)(E_{1}(z+u)-E_{1}(u)),% \qquad f(-z,-u)=f(z,u)}\end{array}start_ARRAY start_ROW start_CELL italic_f ( italic_z , italic_u ) = italic_ฯ• ( italic_z , italic_u ) ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z + italic_u ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u ) ) , italic_f ( - italic_z , - italic_u ) = italic_f ( italic_z , italic_u ) end_CELL end_ROW end_ARRAY (A.4)

in terms of the first Eisenstein function:

E1โข(z)=ฯ‘โ€ฒโข(z)ฯ‘โข(z)=ฮถโข(z)+z3โขฯ‘โ€ฒโ€ฒโ€ฒโข(0)ฯ‘โ€ฒโข(0),E2โข(z)=โˆ’โˆ‚zE1โข(z)=โ„˜โข(z)โˆ’ฯ‘โ€ฒโ€ฒโ€ฒโข(0)3โขฯ‘โ€ฒโข(0),formulae-sequencesubscript๐ธ1๐‘งsuperscriptitalic-ฯ‘โ€ฒ๐‘งitalic-ฯ‘๐‘ง๐œ๐‘ง๐‘ง3superscriptitalic-ฯ‘โ€ฒโ€ฒโ€ฒ0superscriptitalic-ฯ‘โ€ฒ0subscript๐ธ2๐‘งsubscript๐‘งsubscript๐ธ1๐‘งWeierstrass-p๐‘งsuperscriptitalic-ฯ‘โ€ฒโ€ฒโ€ฒ03superscriptitalic-ฯ‘โ€ฒ0\begin{array}[]{c}\displaystyle{E_{1}(z)=\frac{\vartheta^{\prime}(z)}{% \vartheta(z)}=\zeta(z)+\frac{z}{3}\frac{\vartheta^{\prime\prime\prime}(0)}{% \vartheta^{\prime}(0)}\,,\quad E_{2}(z)=-\partial_{z}E_{1}(z)=\wp(z)-\frac{% \vartheta^{\prime\prime\prime}(0)}{3\vartheta^{\prime}(0)}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) = divide start_ARG italic_ฯ‘ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_z ) end_ARG start_ARG italic_ฯ‘ ( italic_z ) end_ARG = italic_ฮถ ( italic_z ) + divide start_ARG italic_z end_ARG start_ARG 3 end_ARG divide start_ARG italic_ฯ‘ start_POSTSUPERSCRIPT โ€ฒ โ€ฒ โ€ฒ end_POSTSUPERSCRIPT ( 0 ) end_ARG start_ARG italic_ฯ‘ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( 0 ) end_ARG , italic_E start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ( italic_z ) = - โˆ‚ start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) = โ„˜ ( italic_z ) - divide start_ARG italic_ฯ‘ start_POSTSUPERSCRIPT โ€ฒ โ€ฒ โ€ฒ end_POSTSUPERSCRIPT ( 0 ) end_ARG start_ARG 3 italic_ฯ‘ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( 0 ) end_ARG , end_CELL end_ROW end_ARRAY (A.5)
E1โข(โˆ’z)=โˆ’E1โข(z),E2โข(โˆ’z)=E2โข(z),formulae-sequencesubscript๐ธ1๐‘งsubscript๐ธ1๐‘งsubscript๐ธ2๐‘งsubscript๐ธ2๐‘ง\begin{array}[]{c}\displaystyle{E_{1}(-z)=-E_{1}(z)\,,\quad E_{2}(-z)=E_{2}(z)% \,,}\end{array}start_ARRAY start_ROW start_CELL italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( - italic_z ) = - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) , italic_E start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ( - italic_z ) = italic_E start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ( italic_z ) , end_CELL end_ROW end_ARRAY (A.6)

where โ„˜โข(z)Weierstrass-p๐‘ง\wp(z)โ„˜ ( italic_z ) and ฮถโข(z)๐œ๐‘ง\zeta(z)italic_ฮถ ( italic_z ) are the Weierstrass functions. We use the following addition formulae:

ฯ•โข(z1,u1)โขฯ•โข(z2,u2)=ฯ•โข(z1,u1+u2)โขฯ•โข(z2โˆ’z1,u2)+ฯ•โข(z2,u1+u2)โขฯ•โข(z1โˆ’z2,u1),italic-ฯ•subscript๐‘ง1subscript๐‘ข1italic-ฯ•subscript๐‘ง2subscript๐‘ข2italic-ฯ•subscript๐‘ง1subscript๐‘ข1subscript๐‘ข2italic-ฯ•subscript๐‘ง2subscript๐‘ง1subscript๐‘ข2italic-ฯ•subscript๐‘ง2subscript๐‘ข1subscript๐‘ข2italic-ฯ•subscript๐‘ง1subscript๐‘ง2subscript๐‘ข1\begin{array}[]{c}\displaystyle{\phi(z_{1},u_{1})\phi(z_{2},u_{2})=\phi(z_{1},% u_{1}+u_{2})\phi(z_{2}-z_{1},u_{2})+\phi(z_{2},u_{1}+u_{2})\phi(z_{1}-z_{2},u_% {1})\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT , italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) = italic_ฯ• ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) + italic_ฯ• ( italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT - italic_z start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) , end_CELL end_ROW end_ARRAY (A.7)
ฯ•โข(z,u1)โขฯ•โข(z,u2)=ฯ•โข(z,u1+u2)โข(E1โข(z)+E1โข(u1)+E1โข(u2)โˆ’E1โข(z+u1+u2)),italic-ฯ•๐‘งsubscript๐‘ข1italic-ฯ•๐‘งsubscript๐‘ข2italic-ฯ•๐‘งsubscript๐‘ข1subscript๐‘ข2subscript๐ธ1๐‘งsubscript๐ธ1subscript๐‘ข1subscript๐ธ1subscript๐‘ข2subscript๐ธ1๐‘งsubscript๐‘ข1subscript๐‘ข2\begin{array}[]{c}\displaystyle{\phi(z,u_{1})\phi(z,u_{2})=\phi(z,u_{1}+u_{2})% \Big{(}E_{1}(z)+E_{1}(u_{1})+E_{1}(u_{2})-E_{1}(z+u_{1}+u_{2})\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) = italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) + italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z + italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (A.8)
ฯ•โข(z,u1)โขfโข(z,u2)โˆ’ฯ•โข(z,u2)โขfโข(z,u1)=ฯ•โข(z,u1+u2)โข(โ„˜โข(u1)โˆ’โ„˜โข(u2)),italic-ฯ•๐‘งsubscript๐‘ข1๐‘“๐‘งsubscript๐‘ข2italic-ฯ•๐‘งsubscript๐‘ข2๐‘“๐‘งsubscript๐‘ข1italic-ฯ•๐‘งsubscript๐‘ข1subscript๐‘ข2Weierstrass-psubscript๐‘ข1Weierstrass-psubscript๐‘ข2\begin{array}[]{c}\displaystyle{\phi(z,u_{1})f(z,u_{2})-\phi(z,u_{2})f(z,u_{1}% )=\phi(z,u_{1}+u_{2})\Big{(}\wp(u_{1})-\wp(u_{2})\Big{)}\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) italic_f ( italic_z , italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) - italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) italic_f ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) = italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ( โ„˜ ( italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) - โ„˜ ( italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ) , end_CELL end_ROW end_ARRAY (A.9)
ฯ•โข(z,u)โขฯ•โข(z,โˆ’u)=โ„˜โข(z)โˆ’โ„˜โข(u)=E2โข(z)โˆ’E2โข(u),italic-ฯ•๐‘ง๐‘ขitalic-ฯ•๐‘ง๐‘ขWeierstrass-p๐‘งWeierstrass-p๐‘ขsubscript๐ธ2๐‘งsubscript๐ธ2๐‘ข\begin{array}[]{c}\displaystyle{\phi(z,u)\phi(z,-u)=\wp(z)-\wp(u)=E_{2}(z)-E_{% 2}(u)\,,}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z , italic_u ) italic_ฯ• ( italic_z , - italic_u ) = โ„˜ ( italic_z ) - โ„˜ ( italic_u ) = italic_E start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ( italic_z ) - italic_E start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ( italic_u ) , end_CELL end_ROW end_ARRAY (A.10)
ฯ•โข(z,u)โขfโข(z,โˆ’u)โˆ’ฯ•โข(z,โˆ’u)โขfโข(z,u)=โ„˜โ€ฒโข(u).italic-ฯ•๐‘ง๐‘ข๐‘“๐‘ง๐‘ขitalic-ฯ•๐‘ง๐‘ข๐‘“๐‘ง๐‘ขsuperscriptWeierstrass-pโ€ฒ๐‘ข\begin{array}[]{c}\displaystyle{\phi(z,u)f(z,-u)-\phi(z,-u)f(z,u)=\wp^{\prime}% (u)\,.}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z , italic_u ) italic_f ( italic_z , - italic_u ) - italic_ฯ• ( italic_z , - italic_u ) italic_f ( italic_z , italic_u ) = โ„˜ start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_u ) . end_CELL end_ROW end_ARRAY (A.11)

The second order partial derivative fโ€ฒโข(z,u)=โˆ‚u2ฯ•โข(z,u)superscript๐‘“โ€ฒ๐‘ง๐‘ขsubscriptsuperscript2๐‘ขitalic-ฯ•๐‘ง๐‘ขf^{\prime}(z,u)=\partial^{2}_{u}\phi(z,u)italic_f start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_z , italic_u ) = โˆ‚ start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_u end_POSTSUBSCRIPT italic_ฯ• ( italic_z , italic_u ):

fโ€ฒโข(z,u)=ฯ•โข(z,u)โข(โ„˜โข(z)โˆ’E12โข(z)+2โขโ„˜โข(u)โˆ’2โขE1โข(z)โขE1โข(u)+2โขE1โข(z+u)โขE1โข(z))==(โ„˜โข(z)โˆ’E12โข(z)+2โขโ„˜โข(u))โขฯ•โข(z,u)+2โขE1โข(z)โขfโข(z,u).superscript๐‘“โ€ฒ๐‘ง๐‘ขitalic-ฯ•๐‘ง๐‘ขWeierstrass-p๐‘งsuperscriptsubscript๐ธ12๐‘ง2Weierstrass-p๐‘ข2subscript๐ธ1๐‘งsubscript๐ธ1๐‘ข2subscript๐ธ1๐‘ง๐‘ขsubscript๐ธ1๐‘งabsentabsentWeierstrass-p๐‘งsuperscriptsubscript๐ธ12๐‘ง2Weierstrass-p๐‘ขitalic-ฯ•๐‘ง๐‘ข2subscript๐ธ1๐‘ง๐‘“๐‘ง๐‘ข\begin{array}[]{c}\displaystyle{f^{\prime}(z,u)=\phi(z,u)\Big{(}\wp(z)-E_{1}^{% 2}(z)+2\wp(u)-2E_{1}(z)E_{1}(u)+2E_{1}(z+u)E_{1}(z)\Big{)}=}\\ \displaystyle{=\Big{(}\wp(z)-E_{1}^{2}(z)+2\wp(u)\Big{)}\phi(z,u)+2E_{1}(z)f(z% ,u)\,.}\end{array}start_ARRAY start_ROW start_CELL italic_f start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_z , italic_u ) = italic_ฯ• ( italic_z , italic_u ) ( โ„˜ ( italic_z ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) + 2 โ„˜ ( italic_u ) - 2 italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u ) + 2 italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z + italic_u ) italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) ) = end_CELL end_ROW start_ROW start_CELL = ( โ„˜ ( italic_z ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) + 2 โ„˜ ( italic_u ) ) italic_ฯ• ( italic_z , italic_u ) + 2 italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_f ( italic_z , italic_u ) . end_CELL end_ROW end_ARRAY (A.12)

Introduce the function

Fโข(z,u)=E1โข(z)โขฯ•โข(z,u)โˆ’fโข(z,u)=(E1โข(z)+E1โข(u)โˆ’E1โข(z+u))โขฯ•โข(z,u).๐น๐‘ง๐‘ขsubscript๐ธ1๐‘งitalic-ฯ•๐‘ง๐‘ข๐‘“๐‘ง๐‘ขsubscript๐ธ1๐‘งsubscript๐ธ1๐‘ขsubscript๐ธ1๐‘ง๐‘ขitalic-ฯ•๐‘ง๐‘ข\begin{array}[]{c}\displaystyle{F(z,u)=E_{1}(z)\phi(z,u)-f(z,u)=\Big{(}E_{1}(z% )+E_{1}(u)-E_{1}(z+u)\Big{)}\phi(z,u)\,.}\end{array}start_ARRAY start_ROW start_CELL italic_F ( italic_z , italic_u ) = italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_ฯ• ( italic_z , italic_u ) - italic_f ( italic_z , italic_u ) = ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) + italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z + italic_u ) ) italic_ฯ• ( italic_z , italic_u ) . end_CELL end_ROW end_ARRAY (A.13)

Then

Fโ€ฒโข(z,u)โ‰กโˆ‚uFโข(z,u)=โˆ’(โ„˜โข(z)โˆ’E12โข(z)+2โขโ„˜โข(u))โขฯ•โข(z,u)โˆ’E1โข(z)โขfโข(z,u)==E1โข(z)โขFโข(z,u)โˆ’(โ„˜โข(z)+2โขโ„˜โข(u))โขฯ•โข(z,u).superscript๐นโ€ฒ๐‘ง๐‘ขsubscript๐‘ข๐น๐‘ง๐‘ขWeierstrass-p๐‘งsuperscriptsubscript๐ธ12๐‘ง2Weierstrass-p๐‘ขitalic-ฯ•๐‘ง๐‘ขsubscript๐ธ1๐‘ง๐‘“๐‘ง๐‘ขabsentabsentsubscript๐ธ1๐‘ง๐น๐‘ง๐‘ขWeierstrass-p๐‘ง2Weierstrass-p๐‘ขitalic-ฯ•๐‘ง๐‘ข\begin{array}[]{c}\displaystyle{F^{\prime}(z,u)\equiv\partial_{u}F(z,u)=-\Big{% (}\wp(z)-E_{1}^{2}(z)+2\wp(u)\Big{)}\phi(z,u)-E_{1}(z)f(z,u)=}\\ \displaystyle{=E_{1}(z)F(z,u)-\Big{(}\wp(z)+2\wp(u)\Big{)}\phi(z,u)\,.}\end{array}start_ARRAY start_ROW start_CELL italic_F start_POSTSUPERSCRIPT โ€ฒ end_POSTSUPERSCRIPT ( italic_z , italic_u ) โ‰ก โˆ‚ start_POSTSUBSCRIPT italic_u end_POSTSUBSCRIPT italic_F ( italic_z , italic_u ) = - ( โ„˜ ( italic_z ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) + 2 โ„˜ ( italic_u ) ) italic_ฯ• ( italic_z , italic_u ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_f ( italic_z , italic_u ) = end_CELL end_ROW start_ROW start_CELL = italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_F ( italic_z , italic_u ) - ( โ„˜ ( italic_z ) + 2 โ„˜ ( italic_u ) ) italic_ฯ• ( italic_z , italic_u ) . end_CELL end_ROW end_ARRAY (A.14)

Also,

ฯ•โข(z,u1)โขfโข(z,u2)=ฯ•โข(z,u1+u2)โข(E12โข(z)โˆ’โ„˜โข(z)โˆ’โ„˜โข(u1+u2)โˆ’โ„˜โข(u2))โˆ’โˆ’E1โข(z)โขfโข(z,u1+u2)+(E1โข(u1)+E1โข(u2)โˆ’E1โข(u1+u2))โขfโข(z,u1+u2)italic-ฯ•๐‘งsubscript๐‘ข1๐‘“๐‘งsubscript๐‘ข2limit-fromitalic-ฯ•๐‘งsubscript๐‘ข1subscript๐‘ข2superscriptsubscript๐ธ12๐‘งWeierstrass-p๐‘งWeierstrass-psubscript๐‘ข1subscript๐‘ข2Weierstrass-psubscript๐‘ข2subscript๐ธ1๐‘ง๐‘“๐‘งsubscript๐‘ข1subscript๐‘ข2subscript๐ธ1subscript๐‘ข1subscript๐ธ1subscript๐‘ข2subscript๐ธ1subscript๐‘ข1subscript๐‘ข2๐‘“๐‘งsubscript๐‘ข1subscript๐‘ข2\begin{array}[]{c}\displaystyle{\phi(z,u_{1})f(z,u_{2})=\phi(z,u_{1}+u_{2})% \Big{(}E_{1}^{2}(z)-\wp(z)-\wp(u_{1}+u_{2})-\wp(u_{2})\Big{)}-}\\ \displaystyle{-E_{1}(z)f(z,u_{1}+u_{2})+\Big{(}E_{1}(u_{1})+E_{1}(u_{2})-E_{1}% (u_{1}+u_{2})\Big{)}f(z,u_{1}+u_{2})}\end{array}start_ARRAY start_ROW start_CELL italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) italic_f ( italic_z , italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) = italic_ฯ• ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_z ) - โ„˜ ( italic_z ) - โ„˜ ( italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) - โ„˜ ( italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ) - end_CELL end_ROW start_ROW start_CELL - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) italic_f ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) + ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) + italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) ) italic_f ( italic_z , italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_u start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) end_CELL end_ROW end_ARRAY (A.15)

and

(ฮถโข(z+w)โˆ’ฮถโข(z)โˆ’ฮถโข(w))2=(E1โข(z+w)โˆ’E1โข(z)โˆ’E1โข(w))2=โ„˜โข(z)+โ„˜โข(w)+โ„˜โข(z+w).superscript๐œ๐‘ง๐‘ค๐œ๐‘ง๐œ๐‘ค2superscriptsubscript๐ธ1๐‘ง๐‘คsubscript๐ธ1๐‘งsubscript๐ธ1๐‘ค2Weierstrass-p๐‘งWeierstrass-p๐‘คWeierstrass-p๐‘ง๐‘ค\begin{array}[]{c}\displaystyle{\Big{(}\zeta(z+w)-\zeta(z)-\zeta(w)\Big{)}^{2}% =\Big{(}E_{1}(z+w)-E_{1}(z)-E_{1}(w)\Big{)}^{2}=\wp(z)+\wp(w)+\wp(z+w)\,.}\end% {array}start_ARRAY start_ROW start_CELL ( italic_ฮถ ( italic_z + italic_w ) - italic_ฮถ ( italic_z ) - italic_ฮถ ( italic_w ) ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = ( italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z + italic_w ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_z ) - italic_E start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_w ) ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT = โ„˜ ( italic_z ) + โ„˜ ( italic_w ) + โ„˜ ( italic_z + italic_w ) . end_CELL end_ROW end_ARRAY (A.16)

Acknowledgments

This work was performed at the Steklov International Mathematical Center and supported by the Ministry of Science and Higher Education of the Russian Federation (agreement no. 075-15-2022-265).

References

  • [1] A.A. Akhmetshin, I.M. Krichever, Y.S. Volvovski, Funct. Anal. Appl., 36:4 (2002) 253โ€“-266;
    arXiv:hep-th/0203192.
  • [2] K. Atalikov, A. Zotov, J. Geom. Phys., 164 (2021) 104161; arXiv:2010.14297 [hep-th]. K. Atalikov, A. Zotov, JETP Letters, 117:8 (2023), 630โ€“634; arXiv:2303.08020 [hep-th]. K. Atalikov, A. Zotov, Theoret. and Math. Phys., 219:3 (2024), 1004โ€“1017; arXiv:2403.00428 [hep-th].
  • [3] K. Atalikov, A. Zotov, JETP Lett. 115, 757-762 (2022); arXiv:2204.12576 [math-ph].
  • [4] E. Billey, J. Avan, O. Babelon, Physics Letters A, 186 (1994) 114โ€“118; hep-th/9312042. J. Avan, O. Babelon, E. Billey, Commun. Math. Phys. 178 (1996) 281โ€“300; arXiv: hep-th/9505091.
  • [5] F. Calogero, Lett. Nuovo Cim. 13 (1975) 411โ€“416. J. Moser, Adv. Math. 16 (1975) 1โ€“23. M.A. Olshanetsky, A.M. Perelomov, Phys. Rep. 71 (1981) 313โ€“400.
  • [6] L.D. Faddeev, L.A. Takhtajan, Hamiltonian methods in the theory of solitons, Springer-Verlag, (1987).
  • [7] J. Gibbons, T. Hermsen, Physica D: Nonlinear Phenomena, 11 (1984) 337โ€“348; S. Wojciechowski, Physics Letters A, 111 (1985) 101โ€“103.
  • [8] I.M. Krichever, Funct. Anal. Appl., 14:4 (1980) 282โ€“290.
  • [9] I. Krichever, Commun. Math. Phys., 229 (2002) 229โ€“-269; arXiv:hep-th/0108110.
  • [10] A. Levin, M. Olshanetsky, A. Zotov, Commun. Math. Phys. 236 (2003) 93โ€“133; arXiv:nlin/0110045. A.V. Zotov, A.V. Smirnov, Theoret. and Math. Phys., 177, (2013) 1281โ€“1338.
  • [11] A. Weil, Elliptic functions according to Eisenstein and Kronecker, Springer-Verlag, (1976).
  • [12] A. Zabrodin, A. Zotov, JHEP 07 (2022) 023; arXiv: 2107.01697 [math-ph].
  • [13] V.E. Zakharov, A.B. Shabat, Soviet physics JETP 34:1 (1972) 62โ€“69. V.E. Zakharov, A.B. Shabat, Funct. Anal. Appl., 8:3 (1974), 226โ€“-235. V.E. Zakharov, A.B. Shabat, Funct. Anal. Appl., 13:3 (1979), 166โ€“-174.
  • [14] A. Zotov, arXiv:2404.01898 [hep-th].
  • [15] A.V. Zotov, 1+1 Gaudin Model, SIGMA 7 (2011), 067; arXiv:1012.1072 [math-ph].