Symmetry, Integrability and Geometry: Methods and Applications SIGMA 7(2011), 008, 16 pages A Vertex Operator Approach for Form Factors Z Z of Belavin’s ( /n )-Symmetric Model and Its Application⋆ Yas-Hiro QUANO Department of Clinical Engineering, Suzuka University of Medical Science, Kishioka-cho, Suzuka 510-0293, Japan E-mail: [email protected] 1 1 Received October 22, 2010, in final form January 07, 2011; Published online January 15, 2011 0 2 doi:10.3842/SIGMA.2011.008 n a Abstract. A vertex operator approach for form factors of Belavin’s (Z/nZ)-symmetric J 5 model is constructed on the basis of bosonization of vertex operators in the A(n1−)1 model and vertex-face transformation. As simple application for n = 2, we obtain expressions for 1 z x 2m-point form factors related to the σ and σ operators in the eight-vertex model. ] h Key words: vertex operator approach; form factors; Belavin’s (Z/nZ)-symmetric model; p integral formulae - h 2010 Mathematics Subject Classification: 37K10;81R12 t a m [ 1 Introduction 5 v 2 In [1] and [2] we derived the integral formulae for correlation functions and form factors, respec- 8 tively, of Belavin’s (Z/nZ)-symmetric model [3, 4] on the basis of vertex operator approach [5]. 3 Belavin’s(Z/nZ)-symmetricmodelisann-stategeneralizationofBaxter’seight-vertexmodel[6], 4 . which has (Z/2Z)-symmetries. As for the eight-vertex model, the integral formulae for correla- 0 1 tion functions and form factors were derived by Lashkevich and Pugai [7] and by Lashkevich [8], 0 respectively. 1 It was found in [7] that the correlation functions of the eight-vertex model can be obtained : v by using the free field realization of the vertex operators in the eight-vertex SOS model [9], with i X insertion of the nonlocal operator Λ, called ‘the tail operator’. The vertex operator approach for r higher spin generalization of the eight-vertex model was presented in [10]. The vertex operator a approach for higher rank generalization was presented in [1]. The expression of the spontaneous polarization of the (Z/nZ)-symmetric model [11] was also reproduced in [1], on the basis of vertex operatorapproach. Concerningformfactors, thebosonization schemefortheeight-vertex model was constructed in [8]. The higher rank generalization of [8] was presented in [2]. It was shown in [12, 13] that the elliptic algebra U (sl ) relevant to the (Z/nZ)-symmetric model q,p N provides the Drinfeld realization of the face type elliptic quantum group (sl ) tensored by q,λ N B a Heisenberg algebra. b The present paper is organized as follows. In Section 2 we review the basbic definitions of the (Z/nZ)-symmetric model [3], the corresponding dual face model A(1) model [14], and the n 1 vertex-face correspondence. In Section 3 we summarize the vertex operat−or algebras relevant to the (Z/nZ)-symmetric model and the A(1) model [1, 2]. In Section 4 we construct the free field n 1 representations of the tail operators, in t−erms of those of the basic operators for the type I [15] ⋆ThispaperisacontributiontotheProceedingsoftheInternationalWorkshop“RecentAdvancesinQuantum Integrable Systems”. The full collection is available at http://www.emis.de/journals/SIGMA/RAQIS2010.html 2 Y.-H. Quano (1) and the type II [16] vertex operators in the A model. Note that in the present paper we n 1 use a different convention from the one used in [−1, 2]. In Section 5 we calculate 2m-point form factors of the σz-operator and σx-operator in the eight-vertex model, as simple application for n = 2. InSection 6wegive someconcludingremarks. Usefuloperator productexpansion(OPE) formulae and commutation relations for basic bosons are given in Appendix A. 2 Basic definitions The present section aims to formulate the problem, thereby fixing the notation. 2.1 Theta functions The Jacobi theta function with two pseudo-periods 1 and τ (Imτ > 0) are defined as follows: a ϑ (v;τ) := exp π√ 1(m+a) [(m+a)τ +2(v+b)] , b − (cid:20) (cid:21) m Z X∈ (cid:8) (cid:9) for a,b R. Let n Z>2 and r R>1, and also fix the parameter x such that 0 < x < 1. We ∈ ∈ ∈ will use the abbreviations, [v] := xvr2−vΘx2r(x2v), [v]′ := [v]r r 1, [v]1 := [v]r 1, | 7→ − | 7→ v := xvr2−vΘx2r( x2v), v ′ := v r r 1, v 1 := v r 1, { } − { } { }| 7→ − { } { }| 7→ where Θ (z) = (z;q) qz 1;q (q;q) = qm(m 1)/2( z)m, q − − ∞ ∞ ∞ m Z − (cid:0) (cid:1) X∈ (z;q ,...,q ) = 1 zqi1 qim . 1 m ∞ − 1 ··· m i1,..Y.,im>0(cid:0) (cid:1) Note that 1/2 v π√ 1 ǫr ǫr ϑ , − = exp [v], 1/2 r ǫr π − 4 (cid:20) − (cid:21)(cid:18) (cid:19) r (cid:16) (cid:17) 0 v π√ 1 ǫr ǫr ϑ , − = exp v , 1/2 r ǫr π − 4 { } (cid:20) (cid:21)(cid:18) (cid:19) r (cid:16) (cid:17) where x = e ǫ (ǫ > 0). − For later conveniences we also introduce the following symbols: r−1n−j gj(z−1) x2n+2r−j−1z xj+1z rj(v) = z r n g (z) , gj(z) = {x2n j+1z x}2{r+j 1z}, (2.1) j − − { }{ } r n−j gj∗(z−1) x2n+2r−j−1z ′ xj−1z ′ rj∗(v) = zr−1 n g (z) , gj∗(z) = {x2n j 1z x}2{r+j 1z} , (2.2) j∗ { − − }′{ − }′ j(n−j) ρj(z−1) (x2j+1z;x2,x2n) (x2n−2j+1z;x2,x2n) χj(v) = (−z)− n ρ (z) , ρj(z) = (xz;x2,x2n)∞(x2n+1z;x2,x2n) ∞, (2.3) j ∞ ∞ where z =x2v, 1 6 j 6 n and z = (z;x2r,x2n) , z = (z;x2r 2,x2n) . ′ − { } ∞ { } ∞ Form Factors of Belavin’s (Z/nZ)-Symmetric Model and Its Application 3 In particular we denote χ(v) = χ (v). These factors will appear in the commutation relations 1 among the type I and type II vertex operators. TheintegralkernelforthetypeIandthetypeIIvertexoperatorswillbegivenastheproducts of the following elliptic functions: [v+ 1 w] [v 1] f(v,w) = 2 − , h(v) = − , [v 1] [v+1] − 2 [v 1 +w] [v+1] f (v,w) = − 2 ′, h (v) = ′. ∗ [v+ 1] ∗ [v 1] 2 ′ − ′ 2.2 Belavin’s (Z/nZ)-symmetric model Let V = Cn and εµ 06µ6n 1 be the standard orthonormal basis with the inner product ε ,ε = δ . Bela{vin}’s (Z/n−Z)-symmetric model [3] is a vertex model on a two-dimensional µ ν µν h i square lattice such that the state variables take the values of (Z/nZ)-spin. The model is L (Z/nZ)-symmetric in a sense that the R-matrix satisfies the following conditions: (i) R(v)ik = 0, unless i+k = j +l, mod n, jl (ii) R(v)i+pk+p = R(v)ik, i,j,k,l,p Z/nZ. j+pl+p jl ∀ ∈ Thedefinitionof theR-matrix intheprincipalregimecan befoundin[2]. ThepresentR-matrix has three parameters v, ǫ and r, which lie in the following region: ǫ >0, r > 1, 0< v < 1. (1) 2.3 The A model n−1 The dual face model of the (Z/nZ)-symmetric model is called the A(1) model. This is a face n 1 model on a two-dimensional squarelattice , the dual lattice of , such−that the state variables ∗ L L take the values of the dual space of Cartan subalgebra h of A(1) : ∗ n 1 − n 1 − h = Cω , ∗ µ µ=0 M where µ 1 n 1 − 1 − ω := ε¯ , ε¯ = ε ε . µ ν µ µ µ − n ν=0 µ=0 X X The weight lattice P and the root lattice Q of A(1) are usually defined. For a h , we set n 1 ∈ ∗ − n 1 − a = a¯ a¯ , a¯ = a+ρ,ε = a+ρ,ε¯ , ρ = ω . µν µ ν µ µ µ µ − h i h i µ=1 X An ordered pair (a,b) h 2 is called admissible if b = a+ε¯ , for a certain µ(0 6 µ 6 n 1). ∗ µ ∈ − c d For (a,b,c,d) h 4, let W v be the Boltzmann weight of the A(1) model for the state ∈ ∗ b a n 1 (cid:20) (cid:12) (cid:21) − c d (cid:12) configuration round a fa(cid:12)ce. Here the four states a, b, c and d are ordered clockwise b a (cid:12) (cid:20) (cid:21) 4 Y.-H. Quano c d from the SE corner. In this model W v = 0 unless the four pairs (a,b),(a,d),(b,c) b a (cid:20) (cid:12) (cid:21) and (d,c) are admissible. Non-zero Boltzmann(cid:12)weights are parametrized in terms of the elliptic (cid:12) theta function of the spectral parameter v. Th(cid:12)e explicit expressions of W can be found in [2]. We consider the so-called Regime III in the model, i.e., 0 < v < 1. 2.4 Vertex-face correspondence Let t(v)a be the intertwining vectors in Cn, whose elements are expressed in terms of theta a ε¯µ functions−. As for the definitions see [2]. Then t(v)a ’s relate the R-matrix of the (Z/nZ)- a ε¯µ − (1) symmetric model in the principal regime and Boltzmann weights W of the A model in the n 1 regime III − c d R(v v )t(v )d t(v )c = t(v )c t(v )bW v v . (2.4) 1− 2 1 a⊗ 2 d 1 b⊗ 2 a b a 1− 2 b (cid:20) (cid:12) (cid:21) X (cid:12) (cid:12) (cid:12) Let us introduce the dual intertwining vectors satisfying n 1 n 1 − t∗µ(v)aa′tµ(v)aa′′ = δaa′′′, − tµ(v)aa ε¯νt∗µ′(v)aa−ε¯ν = δµµ′. (2.5) − µ=0 ν=0 X X From (2.4) and (2.5), we have c d t (v )b t (v )aR(v v ) = W v v t (v )a t (v )d. ∗ 1 c⊗ ∗ 2 b 1− 2 b a 1− 2 ∗ 1 d ⊗ ∗ 2 c d (cid:20) (cid:12) (cid:21) X (cid:12) (cid:12) (cid:12) For fixed r > 1, let c d c d S(v) = R(v) , W v = W v , − |r7→r−1 ′ b a − b a (cid:12) (cid:20) (cid:12) (cid:21) (cid:20) (cid:12) (cid:21) (cid:12)r r 1 (cid:12) (cid:12) (cid:12) 7→ − (cid:12) (cid:12) (cid:12) (cid:12) (cid:12) and t (v)b is the dual intertwining vector of t(v)a. Here, (cid:12) ′∗ a ′ b t(v)a := f (v)t(v;ǫ,r 1)a, ′ b ′ − b with v2 (r+n 2)v (n 1)(3r+n 5) − − − x−n(r 1)− n(r 1) − 6n(r 1) f (v) = − − − ′ n (x2r 2;x2r 2) − − − ∞ (x2z 1;x2n,x2r 2) (x2r+2n 2z;x2n,x2r 2) −p − − − ∞ ∞, (2.6) × (z 1;x2n,x2r 2) (x2r+2n 4z;x2n,x2r 2) − − − − ∞ ∞ and z = x2v. Then we have c d t (v )b t (v )aS(v v )= W v v t (v )a t (v )d. ′∗ 1 c⊗ ′∗ 2 b 1− 2 ′ b a 1− 2 ′∗ 1 d ⊗ ′∗ 2 c d (cid:20) (cid:12) (cid:21) X (cid:12) (cid:12) (cid:12) Form Factors of Belavin’s (Z/nZ)-Symmetric Model and Its Application 5 3 Vertex operator algebra 3.1 Vertex operators for the (Z/nZ)-symmetric model Let (i) be the C-vector space spanned by the half-infinite pure tensor vectors of the forms H ε ε ε with µ Z/nZ, µ = i+1 j (mod n) for j 0. µ1 ⊗ µ2 ⊗ µ3 ⊗··· j ∈ j − ≫ The type I vertex operator Φµ(v) can be defined as a half-infinite transfer matrix. The opera- tor Φµ(v) is an intertwiner from (i) to (i+1), satisfying the following commutation relation: H H Φµ(v )Φν(v )= R(v v )µν Φν′(v )Φµ′(v ). 1 2 1− 2 µ′ν′ 2 1 µ′,ν′ X When we consider an operator related to ‘creation-annihilation’ process, we need another type of vertex operators, the type II vertex operators that satisfy the following commutation relations: Ψ (v )Ψ (v )= Ψ (v )Ψ (v )S(v v )µ′ν′, ∗ν 2 ∗µ 1 ∗µ′ 1 ∗ν′ 2 1 − 2 µν µ′,ν′ X Φµ(v )Ψ (v )= χ(v v )Ψ (v )Φµ(v ). 1 ∗ν 2 1− 2 ∗ν 2 1 Let ρ(i) = x2nHCTM : (i) (i), H → H where H is the CTM Hamiltonian defined as follows: CTM 1 ∞ H (µ ,µ ,µ ,...) = jH (µ ,µ ), CTM 1 2 3 v j j+1 n j=1 X µ ν 1 if 06 ν < µ 6 n 1, Hv(µ,ν) = n−1+−µ ν if 06 µ 6 ν 6 n−1. (3.1) (cid:26) − − − Then we have the homogeneity relations Φµ(v)ρ(i) = ρ(i+1)Φµ(v n), Ψ (v)ρ(i) = ρ(i+1)Ψ (v n). − ∗µ ∗µ − (1) 3.2 Vertex operators for the A model n−1 For k = a+ρ,l = ξ+ρand06 i 6 n 1,let (i) bethespaceofadmissiblepaths(a ,a ,a ,...) − Hl,k 0 1 2 such that a = a, a a ε¯ ,ε¯ ,...,ε¯ for j =0,1,2,3,..., 0 j j+1 0 1 n 1 − ∈ { − } a = ξ+ω for j 0. j i+1 j − ≫ The type I vertex operator Φ(v)a+ε¯µ can be defined as a half-infinite transfer matrix. The a operator Φ(v)a+ε¯µ is an intertwiner from (i) to (i+1) , satisfying the following commutation a Hl,k Hl,k+ε¯µ relation: c d Φ(v )cΦ(v )b = W v v Φ(v )cΦ(v )d. 1 b 2 a b a 1− 2 2 d 1 a d (cid:20) (cid:12) (cid:21) X (cid:12) (cid:12) The free field realization of Φ(v )b w(cid:12)as constructed in [15]. See Section 4.2. 2 a 6 Y.-H. Quano The type II vertex operators should satisfy the following commutation relations: ξ ξ Ψ (v )ξcΨ (v )ξd = Ψ (v )ξcΨ (v )ξbW c d v v , ∗ 2 ξd ∗ 1 ξa Xξb ∗ 1 ξb ∗ 2 ξa ′(cid:20) ξb ξa (cid:12)(cid:12) 1− 2(cid:21) Φ(v1)aa′Ψ∗(v2)ξξ′ = χ(v1−v2)Ψ∗(v2)ξξ′Φ(v1)aa′. (cid:12)(cid:12) Let ρ(l,ik) = Gax2nHl(,ik), Ga = [aµν], 06µ<ν6n 1 Y − (i) (1) where H is the CTM Hamiltonian of A model in regime III is given as follows: l,k n 1 − (i) 1 ∞ H (a ,a ,a ,...)= jH (a ,a ,a ), l,k 0 1 2 n f j−1 j j+1 j=1 X H (a+ε¯ +ε¯ ,a+ε¯ ,a) = H (ν,µ), f µ ν µ v and H (ν,µ) is the same one as (3.1). Then we have the homogeneity relations v (i) (i+1) ρ ρ Φ(v)a′ a+ρ,l = a′+ρ,lΦ(v n)a′, Ψ (v)ξ′ρ(i) = ρ(i+1) Ψ (v n)ξ′. a Ga Ga′ − a ∗ ξ k,ξ+ρ k,ξ′+ρ ∗ − ξ ξ′ The free field realization of Ψ (v) was constructed in [16]. See Section 4.3. ∗ ξ 3.3 Tail operators and commutation relations In [1] we introduced the intertwining operators between (i) and (i) (k = l+ω (mod Q)): H Hl,k i T(u)ξa0 = ∞ tµj( u)aj : (i) (i), − aj+1 H → Hl,k j=0 Y T(u) = ∞ t ( u)aj+1 : (i) (i), ξa0 ∗µj − aj Hl,k → H j=0 Y which satisfy ρ(i) = (x2(rx−22r;;xx22rr−)2)∞ (n−1)(n−2)/2 G1 T(u)aξρ(l,ik)T(u)aξ. (3.2) (cid:18) ∞ (cid:19) ′ξ k≡Xl+ωi (modQ) In order to obtain the form factors of the (Z/nZ)-symmetric model, we need the free field representations ofthetailoperatorwhichisoffdiagonalwithrespecttotheboundaryconditions: Λ(u)ξ′a′ =T(u)ξ′a′T(u) : (i) (i) , (3.3) ξa ξa Hl,k → Hl′k′ where k = a+ρ, l = ξ+ρ, k = a +ρ, and l = ξ +ρ. Let ′ ′ ′ ′ n 1 L a′0 a′1 u := − t ( u)a1tµ( u)a′0. (cid:20) a0 a1 (cid:12) (cid:21) µ=0 ∗µ − a0 − a′1 (cid:12) X (cid:12) (cid:12) Form Factors of Belavin’s (Z/nZ)-Symmetric Model and Its Application 7 Then we have Λ(u)ξ′a′0 = ∞ L a′j a′j+1 u . ξa0 j=0 (cid:20) aj aj+1 (cid:12) (cid:21) Y (cid:12) (cid:12) From the invertibility of the int(cid:12)ertwining vector and its dual vector, we have ξ′a ξ′ Λ(u ) = δ . (3.4) 0 ξa ξ Note that the tail operator (3.3) satisfies the following intertwining relations [1, 2]: Λ(u)ξ′cΦ(v)b = L c d u v Φ(v)cΛ(u)ξ′d, (3.5) ξb a b a − d ξa d (cid:20) (cid:12) (cid:21) X (cid:12) Ψ (v)ξcΛ(u)ξda′ = L ξc(cid:12)(cid:12) ξd u+∆u v Λ(u)ξca′Ψ (v)ξb, (3.6) ∗ ξd ξaa Xξb ′(cid:20) ξb ξa (cid:12)(cid:12) − (cid:21) ξba ∗ ξa (cid:12) where (cid:12) ξ ξ ξ ξ L c d u = L c d u . ′ ξ ξ ξ ξ (cid:20) b a (cid:12) (cid:21) (cid:20) b a (cid:12) (cid:21)(cid:12)r r 1 (cid:12) (cid:12) (cid:12) 7→ − We should find a r(cid:12)(cid:12)epresentation of Λ(cid:12)(cid:12)(u)ξ(cid:12)(cid:12)′a′ and fix the constant ∆u that solves (3.5) and (3.6). ξa 4 Free filed realization 4.1 Bosons In[17,18]thebosonsBj (1 6 j 6 n 1,m Z 0 )relevanttoellipticalgebrawereintroduced. m − ∈ \{ } For α,β h we denote the zero mode operators by P , Q . Concerning commutation relations ∗ α β ∈ among these operators see [17, 18, 2]. We will deal with the bosonic Fock spaces , (l,k h ) generated by Bj (m > 0) over the vacuum vectors l,k : Fl,k ∈ ∗ −m | i Fl,k = C[{Bj1,Bj2,...}16j6n]|l,ki, − − where l,k = exp √ 1(β Q +β Q ) 0,0 , 1 k 2 l | i − | i and (cid:0) (cid:1) 1 t2 β t 1= (t β )(t β ), β = , β < β . 0 1 2 0 1 2 − − − − r(r 1) − p 4.2 Type I vertex operators Let us define the basic operators for j = 1,...,n 1 − j j+1 U−αj(v) = zr−r1 : exp−β1 √−1Qαj +Pαj logz) + Bm−mBm (xjz)−m :, m=0 (cid:0) (cid:1) X6 j Uωj(v) = zr2−r1j(nn−j) :expβ1 √−1Qωj +Pωj logz) − m1 x(j−2k+1)mBmkz−m :, m=0 k=1 (cid:0) (cid:1) X6 X 8 Y.-H. Quano where β = r 1 and z = x2v as usual. The normal product operation places P ’s to the 1 − −r α right of Qβ’s, qas well as Bm’s (m > 0) to the right of B m’s. For some useful OPE formulae − and commutation relations, see Appendix A. In what follows we set π = r(r 1)P , π = π π = rL (r 1)K . µ − ε¯µ µν µ− ν µν − − µν The operatoprs K , L and π act on as scalars ε ε ,k , ε ε ,l and ε ε ,rl µν µν µν l,k µ ν µ ν µ ν F h − i h − i h − − (r 1)k , respectively. In what follows we often use the symbols − i G = [K ], G = [L ]. K µν ′L µν ′ 06µ<ν6n 1 06µ<ν6n 1 Y − Y − For 0 6 µ 6 n 1 the type I vertex operator Φ(v)a+ε¯µ can be expressed in terms of U (v) − a ωj and U (v) on the bosonic Fock space . The explicit expression of Φ(v)a+ε¯µ can found −αj Fl,a+ρ a in [15]. 4.3 Type II vertex operators Let us define the basic operators for j = 1,...,n 1 − j j+1 V−αj(v) = (−z)r−r1 : exp−β2 √−1Qαj +Pαj log(−z) − Am−mAm (xjz)−m:, m=0 (cid:0) (cid:1) X6 r j(n−j) Vωj(v) = (−z)2(r−1) n j 1 :exp β √ 1Q +P log( z) + x(j 2k+1)mAk z m :, × 2 − ωj ωj − m − m − m=0 k=1 (cid:0) (cid:1) X6 X where β = r and z = x2v, and Aj = [rm]x Bj . For some useful OPE formulae and 2 r 1 m [(r 1)m]x m − − commutationqrelations, see Appendix A. For 0 6 µ 6 n 1 the type II vertex operator Ψ (v)ξ+ε¯µ can be expressed in terms of V (v) − ∗ ξ ωj and V (v) on the bosonic Fock space . The explicit expression of Ψ (v)ξ+ε¯µ can found −αj Fξ+ρ,k ∗ ξ in [16]. 4.4 Free field realization of tail operators In order to construct free field realization of the tail operators, we also need another type of basic operators: 1 W−αj(v) = ((−1)rz)r(r−1) j j+1 O O × : exp−β0 √−1Qαj +Pαj log(−1)rz) − m−m m (xjz)−m :, m=0 (cid:0) (cid:1) X6 where β = β + β = 1 , ( 1)r := exp(π√ 1r) and Oj = [m]x Bj . Concerning 0 1 2 √r(r 1) − − m [(r 1)m]x m − − useful OPE formulae and commutation relations, see Appendix A. We cite the results on the free field realization of tail operators. In [1] we obtained the free ξa′ field representation of Λ(u) satisfying (3.5) for ξ = ξ: ξa ′ ν dz Λ(u)ξξaa−−εε¯¯νµ = GKI j=µ+12π√−j1zjU−αµ+1(vµ+1)···U−αν(vν) Y Form Factors of Belavin’s (Z/nZ)-Symmetric Model and Its Application 9 ν 1 − × (−1)Ljν−Kjνf(vj+1−vj,πjν)G−K1, (4.1) j=µ Y where v = u and µ < ν. In [2] we obtained the free field representation of Λ(v)ξ+ε¯n−1a+ε¯n−1 µ ξ+ε¯µa+ε¯n−2 satisfying (3.6) as follows: ( 1)n µ[a ] [ξ 1] Λ(u)ξξ++εε¯¯µn−a+1aε¯+n−ε¯n2−1 = (x−1 −x)(xn2−r;2xn2−r1)3 µn−[11]− ′GKG′L−1 − ′ − ∞ n 2 − dzj W u r 1 V (v ) V (v ) ×IC′j=µ+12π√−1zj −αn−1 − −2 −αn−2 n−2 ··· −αµ+1 µ+1 Y (cid:0) (cid:1) n 2 − × (−1)Lµj−Kµjf∗(vj −vj+1,πµj)G−K1G′L, (4.2) j=µ+1 Y for 0 6 µ 6 n 2 with ∆u = n 1 and v = u. Concerning other types of tail operators ξa′ − − −2 n−1 Λ(u) , the expressions of the free field representation can be found in [1, 2]. ξa 4.5 Free field realization of CTM Hamiltonian Let n 1 j n 1 H = ∞ [rm]x − x(2k 2j 1)mBk (Bj Bj+1)+ 1 − P P F [(r 1)m]x − − −m m− m 2 ωj αj m=1 − j=1k=1 j=1 X XX X n 1 j n 1 = ∞ [rm]x − x(2j 2k 1)m(Bj Bj+1)Bk + 1 − P P (4.3) [(r 1)m]x − − −m− −m m 2 ωj αj m=1 − j=1k=1 j=1 X XX X be the CTM Hamiltonian on the Fock space [19]. Then we have the homogeneity relation l,k F φµ(z)qHF = qHFφµ q−1z , and the trace formula (cid:0) (cid:1) xnβ1k+β2l2 tr x2nHFG = | | G . Fl,k a (x2n;x2n)n−1 a ∞ (cid:0) (cid:1) Let ρ(l,ik) = Gax2nHF. Then the relation (3.2) holds. We thus indentify HF with free field (i) (1) representations of H , the CTM Hamiltonian of A model in regime III. l,k n 1 − 5 Form factors for n = 2 In this section we would like to find explicit expressions of form factors for n = 2 case, i.e., the eight-vertex model form factors. Here, we adopt the convention that the components 0 and 1 for n = 2 are denoted by + and . Form factors of the eight-vertex model are defined as matrix − elements of some local operators. For simplicity, we choose σz as a local operator: σz = E(1) E(1), ++− −− where E(j) is the matrix unit on the j-th site. The free field representation of σz is given by µµ′ σz = εΦ (u)Φε(u). ∗ε ε= X± c 10 Y.-H. Quano Here, Φ (u) is the dual type I vertex operator, whose free filed representation can be found ∗ε in [1, 2]. The corresponding form factors with 2m ‘charged’ particles are given by 1 F(i)(σz;u ,...,u ) = Tr Ψ (u ) Ψ (u )σzρ(i) , (5.1) m 1 2m ν1···ν2m χ(i) H(i) ∗ν1 1 ··· ∗ν2m 2m (cid:16) (cid:17) where c (x4;x4) χ(i) = Tr (i)ρ(i) = (x2;x2)∞. H ∞ In this section we denote the spectral parameters by zj = x2uj, and denote integral variables by w = x2va. a By using the vertex-face transformation, we can rewrite (5.1) as follows: 1 F(i)(σz;u ,...,u ) = t u u + 1 l1 t u u + 1 l2m m 1 2m ν1···ν2m χ(i) ′ν∗1 1− 0 2 l ··· ′∗ν2m 2m− 0 2 l2m−1 l1,X...,l2m (cid:0) (cid:1) (cid:0) (cid:1) ε t (u u )k tε(u u )k1 × ∗ε − 0 k1 − 0 k2 k≡l+Xi(mod2)εX=± k1X=k±1k2=Xk1±1 Tr Ψ (u )l Ψ (u )l2m−1Φ (u)k Φ(u)k1Λ(u )l2mk2[k]x4HF , × (i) ∗ 1 l1··· ∗ 2m l2m ∗ k1 k2 0 lk [l] Hl,k(cid:18) ′ (cid:19) where H is the CTM Hamiltonian defined by (4.3). F Let 1 F(i)(σz;u ,...,u ) = ε t (u u )k tε(u u )k1 m 1 2m ll1···l2m χ(i) ∗ε − 0 k1 − 0 k2 k≡l+Xi(mod2)εX=± k1X=k±1k2=Xk1±1 Tr Ψ (u )l Ψ (u )l2m−1Φ (u)k Φ(u)k1Λ(u )l2mk2[k]x4HF . (5.2) × (i) ∗ 1 l1··· ∗ 2m l2m ∗ k1 k2 0 lk [l] Hl,k(cid:18) ′ (cid:19) Then we have F(i)(σz;u ,...,u ) = F(i)(σz;u ,...,u ) m 1 2m ll1···l2m m 1 2m ν1···ν2m ν1,X...,ν2m tν1 u u + 1 l tν2m u u + 1 l2m−1. × ′ 1− 0 2 l1··· ′ 2m− 0 2 l2m (cid:0) (cid:1) (cid:0) (cid:1) For simplicity, let l =l j for 1 6j 6 2m. Thenfrom therelation (3.4), Λ(u )l2mk2 vanishes j − 0 lk unless k = k 2. Thus, the sum over k and k on (5.2) reduces to only one non-vanishing 2 1 2 − term. Furthermore, we note the formula 0 u u 1+k εt∗ε(u−u0)kk−1tε(u−u0)kk−−21 = (−1)1−i{ }[{u −u00][k− 1] }. ε= − − X± Here, we use k l i (mod 2). The sum with respect to k for the trace over the zero-modes − ≡ parts can be calculated as follows: 2m m 1 u u0 1+k ( zj)2(rr−l1)−k2(x−1z)−l+(r−r1)k − ( wa)−rr−l1+k { − − } − − k l+i(mod2) j=1 a=1 ≡ X Y Y ( 1)rx−r+1z0 −r−l1+kr xrr−l21−2kl+(r−r1)k2 × − (cid:0) (cid:1)