POLYNOMIAL DYNAMICAL SYSTEMS AND DIFFERENTIATION

OF GENUS 4 HYPERELLIPTIC FUNCTIONS

E. YU. BUNKOVA

Abstract. We give an explicit solution to the problem of differentiation of hyperellip-

tic functions in genus 4 case. We describe explicitly the polynomial Lie algebras and
polynomial dynamical systems connected to this problem.

1. Introduction
arXiv:2201.02462v1 [nlin.SI] 7 Jan 2022

Let g ∈ N. We denote the coordinates in the complex space Cg by t = (t1 , t3 , . . . , t2g−1 ).

For a meromorphic function f on Cg , a vector ω ∈ Cg is a period if f (t + ω) = f (t) for
all t ∈ Cg . If a meromorphic function f has 2g independent periods in Cg , then f is called
an Abelian function. Thus, an Abelian function is a meromorphic function on the complex
torus Cg /Γ, where Γ is the lattice formed by the periods. See [1].
A plane nonsingular algebraic curve of genus g determines a lattice Γ as the set of peri-
ods of its holomorphic differentials. The torus Cg /Γ is the Jacobian variety of the curve.
In [2] the problem of differentiation of the field of Abelian functions on Jacobian va-
rieties of genus g curves is considered. In this work we consider a special case of this
problem. Namely, we consider the model of the universal hyperelliptic curve of genus g
Vλ = {(x, y) ∈ C2 : y 2 = x2g+1 + λ4 x2g−1 + λ6 x2g−2 + . . . + λ4g x + λ4g+2 }.
Each curve is defined by specialization of parameters λ = (λ4 , λ6 , . . . , λ4g , λ4g+2 ) ∈ C2g .
The indices of all the coordinates t = (t1 , t3 , . . . , t2g−1 ) ∈ Cg and of the parame-
ters λ = (λ4 , λ6 , . . . , λ4g , λ4g+2 ) ∈ C2g determine their weights. Namely, wt tk = −k
and wt λk = k. For suitable weights of the other variables, all the equations in this paper
are of homogeneous weight.
Let B ⊂ C2g be the subspace of parameters such that the curve Vλ is nonsingular
for λ ∈ B. Then we have B = C2g \Σ, where Σ is the discriminant hypersurface of the
universal curve. In Section 2 we describe the polynomial vector fields in B tangent to Σ.
For each λ ∈ B the set of periods of holomorphic differentials on the curve Vλ generates
a lattice Γλ of rank 2g in Cg . A hyperelliptic function of genus g is a meromorphic function
on Cg × B such that, for each λ ∈ B, its restriction to Cg × λ is an Abelian function with
lattice of periods Γλ . Thus, a hyperelliptic function is a function defined on an open
dense subset of the total space U of the fiber bundle π : U → B with fiber over λ ∈ B
the Jacobian variety Jλ = Cg /Γλ of the curve Vλ . The universal bundle of Jacobians
of hyperelliptic curves U is introduced in [3]. We denote by F the field of hyperelliptic
functions of genus g.
The problem of differentiation of hyperelliptic functions is a genus g analogue of a genus 1
result of [4], see also §1.2 in [2]. We consider this problem in the form:
Problem 1.1 (Problem 1.1 in [5]).
(1) Find the 3g generators of the F -module Der F of derivations of the field F .
(2) Describe the structure of Lie algebra Der F (i.e. find the commutation relations).
A general approach to the solution of this problem is given in [2]. An overview with
examples in the cases of genus g = 1 elliptic and g = 2 hyperelliptic curves is given in [6].
1
 We use the theory of hyperelliptic Kleinian functions (see [7], [8], [9], [10], and [11] for
elliptic functions). Take coordinates (t, λ) in Cg ×B ⊂ C3g . Let σ(t, λ) be the hyperelliptic
sigma function. We denote ∂k = ∂t∂k . We use the notation
ζk = ∂k ln σ(t, λ), ℘k1 ,...,kn = −∂k1 · · · ∂kn ln σ(t, λ),
where n > 2 and ks ∈ {1, 3, . . . , 2g − 1}. The functions ℘k1 ,...,kn give examples of hyper-
elliptic functions.
ϕ
Consider the diagram U ❴ ❴ ❴/ C3g (1)
π ρ
  
B  / C2g
The fiber bundle π : U → B and the embedding B ⊂ C2g are described above. In Section 3
we describe the maps ϕ and ρ following [5]. By Dubrovin–Novikov theorem [3], the
space U is birationally isomorphic to the complex linear space C3g . We construct such
an isomorphism ϕ explicitly. We use a fundamental result from the theory of hyperelliptic
Abelian functions (see [10], Chapter 5): Any hyperelliptic function can be represented
as a rational function in ℘1,k and ℘1,1,k , where k ∈ {1, 3, . . . , 2g − 1}. Theorem 3.1 in
Section 3 gives a set of relations between the derivatives of these functions. We use it to
introduce a set of generators in F . The map ϕ will be determined by this set of generators.
The map ρ will be a polynomial map that makes the diagram (1) commutative.
We denote the ring of polynomials in λ ∈ C2g by P. For the polynomial map ρ : C3g →
C we call a vector field D in C3g projectable if there exists a vector field L in C2g such
2g

that D(ρ∗ f ) = ρ∗ L(f ) for any f ∈ P. The vector field L is the pushforward of D. A
corollary of this definition is that for a projectable vector field D we have D(ρ∗ P) ⊂ ρ∗ P.
We consider the problem:
Problem 1.2. Find 3g polynomial vector fields in C3g projectable for ρ : C3g → C2g
and independent at any point in ρ−1 (B). Construct their polynomial Lie algebra.
There is a direct relation (2) between Problems 1.1 and 1.2. In Section 4 we give
vector fields that give part of both solutions for general genus g. The pushforwards
of these solutions are the polynomial vector fields described in Section 2. In Section 5 we
give polynomial dynamical systems that arise from this relation.
In the case of genus g = 2 the solution to Problem 1.2 and the corresponding homo-
geneous polynomial dynamical systems in C6 are given in Section 4 of [12]. An explicit
solution to Problem 1.1 is given in Theorems B.3 and B.6 of [12]. We recall these results
in Section 6.
In the case of genus g = 3 an explicit solution to Problem 1.1 is given in [5]. The proof
is based on a solution to Problem 1.2. We give these results in Section 7. The corre-
sponding polynomial dynamical systems in C9 are presented in Section 8 of [13].
The case of genus g = 4 is considered in Sections 8, 9, 10, and 11 of this work.
We find the solution to Problem 1.1 in Theorem 8.2 and the solution to Problem 1.2
in Theorem 9.1. Some of the results of these Sections were first obtained in [14]. Here
we complete the descriptions of the Lie algebras, see Section 10. Our approach is based
on a construction of the generators of Der F based on a result from [15].
In all the cases that we consider the generators Lk of Der F , that solve Problem 1.1,
and the polynomial vector fields Dk in C3g , that solve Problem 1.2, are related as
Lk (ϕ∗ bi,j ) = ϕ∗ Dk (bi,j ) (2)
3g
for coordinate functions bi,j ∈ C . We note that these solutions give explicit solutions
of the Problem considered in [2] for hyperelliptic curves Vλ of genus 2, 3, and 4.
2
 2. Polynomial vector fields tangent to the discriminant hypersurface
of the universal curve
Let us define the polynomial Lie algebra of vector fields tangent to the discriminant
hypersurface Σ in C2g . We denote it by LB . Here C2g is the complex linear space with
coordinates (λ), P is the ring of polynomials in (λ) and Σ is defined in Section 1.
For the polynomial Lie algebra LB , the generators {L0 , L2 , L4 , . . . , L4g−2 } are the vector
fields
2g+1
X ∂
L2k = v2k+2,2s−2 (λ) , v2k+2,2s−2 (λ) ∈ P.
s=2
∂λ 2k

By [16], the structure of a polynomial Lie algebra as a P-module with generators

1, L0 , L2 , L4 , . . . , L4g−2 is determined by the polynomial matrices V (λ) = (v2i,2j (λ)),
where i, j = 1, . . . , 2g, and C(λ) = (c2k
2i,2j (λ)), where i, j, k = 0, . . . , 2g − 1, such that
2g−1
X
[L2i , L2j ] = c2k
2i,2j (λ)L2k , [L2i , λ2j+4 ] = v2i+2,2j+2 (λ), [λ2i+4 , λ2j+4] = 0. (3)
k=0

In the case of the Lie algebra LB , explicit expressions for the matrix V (λ) can be found
in Section 4.1 of [17]. The elements of this matrix are given by the following formulas.
For convenience, we assume that λs = 0 for all s ∈ / {0, 4, 6, . . . , 4g, 4g + 2} and λ0 = 1.
Let k, m ∈ {1, 2, . . . , 2g}. If k 6 m, then we set
k−1
X 2k(2g − m + 1)
v2k,2m (λ) = 2(k + m − 2s)λ2s λ2(k+m−s) − λ2k λ2m , (4)
s=0
2g + 1
and if k > m, then we set v2k,2m (λ) = v2m,2k (λ).
The vector field L0 is the Euler vector field; namely, since wt λ2k = 2k, we have
[L0 , λ2k ] = 2kλ2k , [L0 , L2k ] = 2kL2k .
This determines the weights of the vector fields Lk , namely, wt L2k = 2k.
Lemma 2.1 (Lemma 4.3 in [15]).
4(2g − k)
[L2 , L2k ] = 2(k − 1)L2k+2 + (λ2k+2 L0 − λ4 L2k−2 ) .
(2g + 1)
This Lemma determines the polynomials c2k 2,2j (λ) in (3). For general i the polynomi-
2k
als c2i,2j (λ) in (3) are described in Theorem 2.5 of [18]. We can derive them directly from
the explicit expressions (4).
Example 2.2. In the case of genus g = 2 we have
8 8 4 4
[L2 , L4 ] = λ6 L0 − λ4 L2 + 2L6 , [L2 , L6 ] = λ8 L0 − λ4 L4 ,
5 5 5 5
6 6
[L4 , L6 ] = −2λ10 L0 + λ8 L2 − λ6 L4 + 2λ4 L6 .
5 5
Example 2.3. In the case of genus g = 3 the expressions for c2k 2i,2j are given in
Lemma 4.3 of [5]. In particular, we have:
16 12
[L2 , L4 ] = 2L6 + (λ6 L0 − λ4 L2 ) , [L2 , L6 ] = 4L8 + (λ8 L0 − λ4 L4 ) , (5)
7 7
8
[L2 , L8 ] = 6L10 + (λ10 L0 − λ4 L6 ) .
7
In the case of genus g = 4 the expressions for c2k 2,2j are given in Example 8.1.
3
 3. Polynomial map related to
the universal bundle of Jacobians of hyperelliptic curves
Theorem 3.1 ([8], §3 in [9]). For i, k ∈ {1, 3, . . . , 2g − 1} we have the relations
℘1,1,1,i =6℘1,1 ℘1,i + 6℘1,i+2 − 2℘3,i + 2λ4 δi,1 ,
℘1,1,i ℘1,1,k =4 (℘1,1 ℘1,i ℘1,k + ℘1,k ℘1,i+2 + ℘1,i ℘1,k+2 + ℘k+2,i+2) −
− 2(℘1,i ℘3,k + ℘1,k ℘3,i + ℘k,i+4 + ℘i,k+4)+
+ 2λ4 (δi,1 ℘1,k + δk,1 ℘1,i ) + 2λi+k+4(2δi,k + δk,i−2 + δi,k−2 ).
g(g+9)
Corollary 3.2 (Corollary 5.2 in [5]). Consider the map ϕ e : U 99K C 2 with the coor-
g(g+9)
dinates (b, p, λ) in C 2 . Here b = (bi,j ) ∈ C3g with i ∈ {1, 2, 3}, j ∈ {1, 3, . . . , 2g − 1},
g(g−1)
p = (pk,l ) ∈ C 2 with k, l ∈ {3, 5, . . . , 2g − 1} for k 6 l, and λ = (λs ) ∈ C2g
with s ∈ {4, 6, . . . , 4g, 4g + 2}. Set
ϕ
e : (t, λ) 7→ (b1,j , b2,j , b3,j , pk,l , λs ) = (℘1,j (t, λ), ℘1,1,j (t, λ), ℘1,1,1,j (t, λ), 2℘k,l (t, λ), λs ).
g(g+9)
We denote pl,k = pk,l for k, l ∈ {3, 5, . . . , 2g −1}. Then the image of ϕ
e lies in S ⊂ C 2 ,
where S is determined by the set of g(g + 3)/2 equations
b3,1 =6b21,1 + 4b1,3 + 2λ4 ,
b3,k =6b1,1 b1,k + 6b1,k+2 − p3,k ,
b22,1 =4b31,1 + 4b1,1 b1,3 − 4b1,5 + 2p3,3 + 4λ4 b1,1 + 4λ6 ,
b2,1 b2,k =4b21,1 b1,k + 2b1,3 b1,k + 4b1,1 b1,k+2 − 2b1,k+4 −
− b1,1 p3,k + 2p3,k+2 − p5,k + 2λ4 b1,k + 2λ8 δ3,k ,
b2,j b2,k =4b1,1 b1,j b1,k + 4b1,k b1,j+2 + 4b1,j b1,k+2 + 2pk+2,j+2−
− b1,j p3,k − b1,k p3,j − pk,j+4 − pj,k+4 + 2λj+k+4(2δj,k + δk,j−2 + δj,k−2),
for j, k ∈ {3, . . . , 2g − 1} and any variable equal to zero if the index is out of range.
g(g+9)
Theorem 3.3 (Theorem 5.3 in [5]). The projection π1 : C 2 → C3g on the first 3g co-
ordinates gives the isomorphism S ≃ C3g . Therefore, the coordinates (bi,j ) uniformize S.
e : U 99K S ≃ C3g and we denote by ϕ the composition π1 ◦ ϕ
We have ϕ e : U 99K C3g .
We obtain the diagram g(g+9) g(g−1)
C; O 2 C3g × C 2 × C2g (6)
e
ϕ
✇ ❧❧❧
✇
✇
❧❧❧❧❧
❧❧
✇
✇ ϕ ? v ❧❧ π1
❧
U ❴ ❴ ❴/ S ≃ C3g
π3
π ρ
   y
B / C2g
g(g+9)
Corollary 3.4 (Corollary 5.5 in [5]). The projection π3 : C 2 → C2g on the last 2g
coordinates in Corollary 3.2 restricted to S ≃ C3g gives the polynomial map ρ : C3g → C2g .
g(g+9) g(g−1)
Corollary 3.5 (Corollary 5.4 in [5]). The projection π2 : C 2 → C 2 on the
middle g(g − 1)/2 coordinates in Corollary 3.2 restricted to S ≃ C3g gives a polynomial
g(g−1)
map C3g → C 2 .
The proof of these Corollaries in [5] allows to obtain recursive expressions for the
polynomials λs and pk,l in bi,j . We will describe the polynomial maps ρ : C3g → C2g
g(g−1)
and C3g → C 2 explicitly for any g in our next work.
4
 4. Explicit description of polynomial vector fields projectable for ρ
Now let us describe explicitly some of the polynomial vector fields Dk projectable
for ρ : C3g → C2g . We construct them using the relation
Lk (ϕ∗ bi,j ) = ϕ∗ Dk (bi,j ) (7)
for Lk ∈ Der F .
Remark 4.1 (Section 2 in [5]). The operators L2k−1 = ∂2k−1 for k ∈ {1, 2, 3, . . . , g}
belong to the Lie algebra of derivations of F . Their pushforwards for π are zero.
Lemma 4.2 (Lemma 6.2 in [5]). For the polynomial vector field
X ∂ ∂ ∂
D1 = b2,j + b3,j + 4(2b1,1 b2,j + b2,1 b1,j + b2,j+2 )
j
∂b1,j ∂b2,j ∂b3,j
where b2,2g+1 = 0 we have L1 (ϕ∗ bi,j ) = ϕ∗ D1 (bi,j ) for all i ∈ {1, 2, 3}, j ∈ {1, 3, . . . , 2g−1}.
Lemma 4.3 (cf. Lemma 6.3 in [5]). Set ps,1 = 2b1,s . For s ∈ {3, 5, . . . , 2g − 1} for the
polynomial vector fields
g  
1X ∂ ∂ ∂
Ds = D1 (ps,2k−1 ) + D1 (D1 (ps,2k−1 )) + D1 (D1 (D1 (ps,2k−1)))
2 k=1 ∂b1,2k−1 ∂b2,2k−1 ∂b3,2k−1
we have Ls (ϕ∗ bi,j ) = ϕ∗ Ds (bi,j ) for all i ∈ {1, 2, 3}, j ∈ {1, 3, . . . , 2g − 1}.
Corollary 4.4. The polynomial vector fields Ds for s ∈ {1, 3, . . . , 2g−1} are projectable
for the polynomial map ρ : C3g → C2g . Their pushforwards are zero.
Theorem 4.5 (Theorem 6.1 in [13]). The operators
g
X
L0 = L0 − (2s − 1)t2s−1 ∂2s−1 ,
s=1
g−1 g−1
X 4 X
L2 = L2 − ζ1 ∂1 − (2s − 1)t2s−1 ∂2s+1 + λ4 (g − s)t2s+1 ∂2s−1 ,
s=1
2g + 1 s=1
g−2
X
L4 = L4 − ζ3 ∂1 − ζ1 ∂3 − (2s − 1)t2s−1 ∂2s+3 −
s=1
g−1 g−1
X 6 X
− λ4 (2s − 1)t2s+1 ∂2s+1 + λ6 (g − s)t2s+1 ∂2s−1
s=1
2g + 1 s=1
belong to the Lie algebra of derivations of F .
Lemma 4.6 (Equation (22) in [5]). For the polynomial vector field
X ∂ ∂ ∂
D0 = (j + 1)b1,j + (j + 2)b2,j + (j + 3)b3,j .
j
∂b1,j ∂b2,j ∂b3,j
we have L0 (ϕ∗ bi,j ) = ϕ∗ D0 (bi,j ) for all i ∈ {1, 2, 3}, j ∈ {1, 3, . . . , 2g − 1}.
Corollary 4.7. The polynomial vector field D0 is projectable for the polynomial map ρ
with pushforward L0 .
Problem 4.8. Describe explicitly the polynomial vector fields D2 and D4 such that
Lk (ϕ∗ bi,j ) = ϕ∗ Dk (bi,j ) for k = 2, 4 and i ∈ {1, 2, 3}, j ∈ {1, 3, . . . , 2g − 1}.
We will give a solution of this problem based on results of [13] and [2] in our next work.
5
 5. Polynomial dynamical systems related to differentiations
of hyperelliptic functions
In Section 4 the generators Lk of Der F and the polynomial vector fields Dk in C3g
are related by (7). In this Section we give the graded homogeneous polynomial dynamical
systems in C3g determined by these vector fields. We follow the approach of [12], where
such a description is given in the case of genus g = 1 and g = 2. See also Section 8 of [13].
By definition, the dynamical system Sk corresponding to the vector field Dk is given by
∂
bi,j = Dk (bi,j ).
∂τk
The dynamical system S0 corresponding to the Euler vector field D0 is given by
∂
bi,j = (i + j)bi,j , for i ∈ {1, 2, 3}, j ∈ {1, 3, . . . , 2g − 1}.
∂τ0
The dynamical system S1 corresponding to the vector field D1 for j ∈ {1, 3, . . . , 2g − 1}
is given by
∂ ∂
bi,j = bi+1,j , for i ∈ {1, 2}, b3,j = 4(2b1,1 b2,j + b2,1 b1,j + b2,j+2 ),
∂τ1 ∂τ1
where b2,2g+1 = 0.
The dynamical systems Ss for s ∈ {3, 5, . . . , 2g − 1} corresponding to vector fields Ds
for j ∈ {1, 3, . . . , 2g − 1} are given by
∂ 1 ∂ 1 ∂ 1
b1,j = D1 (ps,j ), b2,j = D1 (D1 (ps,j )), b3,j = D1 (D1 (D1 (ps,j ))),
∂τs 2 ∂τs 2 ∂τs 2
where ps,1 = 2b1,s .

6. Lie algebra of derivations of hyperelliptic functions: genus 2 case

The following results were obtained in [12]. We give them in the notation of this work
using the explicit formulas from Section 4.
Theorem 6.1 (Theorem B.3 and Theorem B.6 in [12]). In the case of genus g = 2
the following vector fields give a solution to Problem 1.1
6
L1 = ∂1 , L0 = L0 − t1 ∂1 − 3t3 ∂3 , L4 = L4 − ζ3 ∂1 − ζ1 ∂3 − λ4 t3 ∂3 + λ6 t3 ∂1 ,
5
4 3
L3 = ∂3 , L2 = L2 − ζ1 ∂1 − t1 ∂3 + λ4 t3 ∂1 , L6 = L6 − ζ3 ∂3 + λ8 t3 ∂1 .
5 5
For m, l ∈ {1, 2}, m 6 l, and k ∈ {1, 2, 3, 4, 6}, the commutation relations are
[L0 , Lk ] = kLk , [L1 , L3 ] = 0,
 
4
[L1 , L2 ] = ℘1,1 L1 − L3 , [L3 , L2 ] = ℘1,3 + λ4 L 1 ,
5
 
6
[L1 , L4 ] = ℘1,3 L1 + ℘1,1 L3 , [L3 , L4 ] = ℘3,3 + λ6 L1 + (℘1,3 − λ4 ) L3 ,
5
3
[L1 , L6 ] = ℘1,3 L3 , [L3 , L6 ] = λ8 L1 + ℘3,3 L3 ,
5
X3
1 1
[L2m , L2l+2 ] = c2s
2m,2l+2 (λ)L2s − ℘2m−1,2l−1,3 L1 + ℘1,2m−1,2l−1 L3 .
s=0
2 2
6
 The expressions for the coordinates (λ) and (p) are (see eq. (4.3)-(4.6) and (4.8) in [12]):
1
λ4 = −3b21,1 + b3,1 − 2b1,3 ,
2
1 1 1
λ6 = 2b31,1 + b22,1 − b1,1 b3,1 − 2b1,1 b1,3 + b3,3 ,
4 2 2

1
λ8 = 4b21,1 + b1,3 b1,3 − (b3,1 b1,3 − b2,1 b2,3 + b1,1 b3,3 ),
2
1 1
λ10 = 2b1,1 b21,3 + b22,3 − b1,3 b3,3 , p3,3 = 6b1,1 b1,3 − b3,3 .
4 2
Theorem 6.2 (Section 4 in [12]). In the case of genus g = 2 the homogeneous polyno-
mial vector fields D0 , D1 , D2 , D3 , D4 , D6 give a solution to Problem 1.2. Here the vector
fields D1 , D3 and D0 are determined by Lemmas 4.2, 4.3 and 4.6 for g = 2. Set p1,k = 2b1,k
and qi,j,k = Di (pj,k ) for i, j, k ∈ {1, 3}. The polynomial vector fields D2 , D4 , D6 are deter-
mined by the conditions
8
D2 (b1,1 ) = λ4 + 2b21,1 + 4b1,3 , (8)
5
2
D4 (b1,1 ) = λ6 − 2b1,1 b1,3 + b3,3 ,
5
1 1
D6 (b1,1 ) = λ8 + (b3,1 b1,3 − b1,1 b3,3 ) − b21,3 ,
5 2
4
D2 (b1,3 ) = − λ4 b1,1 + 2b1,1 b1,3 ,
5
6 1
D4 (b1,3 ) = − λ6 b1,1 + 2λ4 b1,3 − 4b21,1 b1,3 + b3,1 b1,3 − b2,1 b2,3 ,
5 2
8
D6 (b1,3 ) = − λ8 b1,1 + 2λ6 b1,3 − 2b1,1 b21,3 − b22,3 + b1,3 b3,3 .
5
and the relations
[D1 , D2 ] = b1,1 D1 − D3 , [D1 , D4 ] = b1,3 D1 + b1,1 D3 , [D1 , D6 ] = b1,3 D3 . (9)
These relations determine the polynomials Dk (bi,j ) for k ∈ {2, 4, 6}, i ∈ {2, 3}, j ∈ {1, 3}.
For m, l ∈ {1, 2}, m 6 l, and k ∈ {1, 2, 3, 4, 6}, the commutation relations are (9) and
[D0 , Dk ] = kDk [D1 , D3 ] = 0,
   
4 1 6
[D3 , D2 ] = b1,3 + λ4 D1 , [D3 , D4 ] = p3,3 + λ6 D1 + (b1,3 − λ4 ) D3 ,
5 2 5
3 1
[D3 , D6 ] = λ8 D1 + p3,3 D3 ,
5 2
X3
1 1
[D2m , D2l+2 ] = c2k
2m,2l+2 (λ)D2k − q2m−1,2l−1,3 D1 + q1,2m−1,2l−1 D3 .
4 4
k=0

Proof. This Theorem follows from Theorem 4.6 in [12] and the expressions in the proof
of this Theorem. Note that this expressions are based on the relation (7) for Lk given in
Theorem 6.1. See also Theorems 4.7 and B.3 in [12] and Section 8 in [5]. 
Let us note that the result of Theorem 6.2 is equivalent to the corresponding result
in Example 15 in [2]. Some misprints in [2] were corrected in [12]. In terms of the work [2],
the polynomial vector fields Dk determine the action of the operators Lk on F . The latter
is determined by the values Lk (℘1,1 ) ∈ F , that correspond to Dk (b1,1 ), given by (8).
7
 7. Lie algebra of derivations of hyperelliptic functions: genus 3 case
Theorem 7.1 (Theorem 10.1 and Corollary 10.2 in [5]). In the case of genus g = 3
the following vector fields give a solution to Problem 1.1
L1 = ∂1 , L3 = ∂3 , L5 = ∂5 ,
L0 = L0 − t1 ∂1 − 3t3 ∂3 − 5t5 ∂5 ,
   
8 4
L2 = L2 − ζ1 − λ4 t3 ∂1 − t1 − λ4 t5 ∂3 − 3t3 ∂5 ,
7 7
   
12 6
L4 = L4 − ζ3 − λ6 t3 ∂1 − ζ1 + λ4 t3 − λ6 t5 ∂3 − (t1 + 3λ4 t5 )∂5 ,
7 7
   
9 8
L6 = L6 − ζ5 − λ8 t3 ∂1 − ζ3 − λ8 t5 ∂3 − (ζ1 + λ4 t3 + 2λ6 t5 ) ∂5 ,
7 7
   
6 10
L8 = L8 + λ10 t3 − λ12 t5 ∂1 − ζ5 − λ10 t5 ∂3 − (ζ3 + λ8 t5 ) ∂5 ,
7 7
 
3 5
L10 = L10 + λ12 t3 − 2λ14 t5 ∂1 + λ12 t5 ∂3 − ζ5 ∂5 ,
7 7
The commutation relations are
[L0 , Lk ] = kLk , k = 1, 2, 3, 4, 5, 6, 8, 10, [L1 , L3 ] = 0, [L1 , L5 ] = 0, [L3 , L5 ] = 0,
   
[L1 , L2] ℘1,1 −1 0  
 [L1 , L4]  ℘1,3 ℘1,1 −1  L1
   
 [L1 , L6]  = ℘1,5 ℘1,3 ℘1,1  L3  ,
   
 [L1 , L8]   0 ℘1,5 ℘1,3  L5
[L1 , L10 ] 0 0 ℘1,5
     
[L3 , L2 ] 5λ4 ℘1,3 − λ4 0 −3  
 [L3 , L4 ]   4λ6   ℘3,3 ℘ − λ 0  L1
     1,3 4 
 [L3 , L6 ]  = 3  3λ8  L1 +  ℘3,5 ℘3,3 ℘1,3 − λ4   
  7    L3 ,
 [L3 , L8 ]  2λ10   0 ℘3,5 ℘3,3  L5
[L3 , L10 ] λ12 0 0 ℘3,5
       
[L5 , L2 ] 2λ4 0 ℘1,5 0 0  
 [L5 , L4 ]   3λ6  3λ4   ℘3,5 ℘ 0  L1
       1,5 
 [L5 , L6 ]  = 2  4λ8  L3 − 2λ6  L5 +  ℘5,5 ℘ ℘   L3  ,
       3,5 1,5 
 [L5 , L8 ]  7 5λ10   λ8   −λ12 ℘5,5 ℘3,5  L5
[L5 , L10 ] 6λ12 0 −2λ14 −λ12 ℘5,5
   5 P   
[L2 , L4 ] c2k
2,4 (λ)L2k −℘1,1,3 ℘1,1,1 0  
P5k=0 2k
 [L2 , L6 ]      L1
  =  P5k=0 c2,6 (λ)L2k  + 1 −℘1,3,3 − ℘1,1,5 ℘1,1,3 ℘1,1,1  L3  , (10)
 [L2 , L8 ]   c2k  2  −2℘1,3,5 ℘1,1,5 ℘1,1,3 
P5k=0 2k 2,8 (λ)L2k L5
[L2 , L10 ] −℘1,5,5 0 ℘1,1,5
k=0 c2,10 (λ)L2k
 P5 
  c2k (λ)L2k  
[L4 , L6 ] k=0 4,6 −℘3,3,3 ℘1,3,3 − 2℘1,1,5 2℘1,1,3
 P5 c2k (λ)L 
 [L4 , L8 ]  P k=0 4,8 2k  −2℘3,3,5 0 2℘1,3,3   
   5  1  L
2k
[L4 , L10 ]  Pk=0 c4,10 (λ)L2k   −℘3,5,5 −℘1,5,5 2℘1,3,5   1 
 = 5  +   L3 .
 [L6 , L8 ]   k=0 c2k (λ)L2k  2 −2℘3,5,5 2℘1,5,5 − ℘3,3,5 ℘3,3,3 
[L , L ] P5 6,8
2k
  −℘ −℘ ℘ + ℘  L5
6 10  k=0 c6,10 (λ)L2k  5,5,5 3,5,5 3,3,5 1,5,5
[L8 , L10 ] P5 2k 0 −℘5,5,5 ℘3,5,5
k=0 c8,10 (λ)L2k
8
 The expressions for the coordinates (λ) and (p) are (see Section 9 in [5]):
1 1 1 1
λ4 = −3b21,1 + b3,1 − 2b1,3 , λ6 = 2b31,1 + b22,1 − b1,1 b3,1 − 2b1,1 b1,3 + b3,3 − 2b1,5 ,
2 4 2 2
1 1
λ8 = 4b21,1 b1,3 − (b3,1 b1,3 − b2,1 b2,3 + b1,1 b3,3 ) + b21,3 − 2b1,1 b1,5 + b3,5 ,
2 2
1 1 1

λ10 = 2b1,1 b21,3 + b22,3 − b1,3 b3,3 − (b3,1 b1,5 − b2,1 b2,5 + b1,1 b3,5 ) + 4b21,1 + 2b1,3 b1,5 ,
4 2 2
1
λ12 = 4b1,1 b1,3 b1,5 − (b3,3 b1,5 − b2,3 b2,5 + b1,3 b3,5 ) + b21,5 ,
2
2 1 2 1
λ14 = 2b1,1 b1,5 + b2,5 − b1,5 b3,5 , p3,3 = 6b1,1 b1,3 − b3,3 + 6b1,5 ,
4 2

p3,5 = 6b1,1 b1,5 − b3,5 , p5,5 = (b3,1 b1,5 − b2,1 b2,5 + b1,1 b3,5 ) − 2 4b21,1 + b1,3 b1,5 .
Theorem 7.2. In the case of genus g = 3 the homogeneous polynomial vector fields Ds
for s ∈ {0, 1, 2, 3, 4, 5, 6, 8, 10} give a solution to Problem 1.2. The vector fields D1 , D3 , D5
and D0 are determined by Lemmas 4.2, 4.3 and 4.6 for g = 3. We set p1,k = 2b1,k
and qi,j,k = Di (pj,k ) for i, j, k ∈ {1, 3, 5}. The polynomial vector fields D2 and D4 are
determined by the conditions
12 4
D2 (b1,1 ) = λ4 + 2b21,1 + 4b1,3 , D4 (b1,1 ) = λ6 − 2b1,1 b1,3 + b3,3 + 2b1,5 ,
7 7
8
D2 (b1,3 ) = − λ4 b1,1 + 2b1,1 b1,3 + 6b1,5 ,
7
12 1
D4 (b1,3 ) = − λ6 b1,1 − 10b21,1 b1,3 + 2b3,1 b1,3 − 4b21,3 − b2,1 b2,3 + 2b1,1 b1,5 ,
7 2
4
D2 (b1,5 ) = − λ4 b1,3 + 2b1,1 b1,5 ,
7
6 1
D4 (b1,5 ) = − λ6 b1,3 − 16b21,1 b1,5 + 3b3,1 b1,5 − 8b1,3 b1,5 − b2,1 b2,5 ,
7 2
and the relations  
    D1
[D1 , D2 ] b −1 0  
= 1,1 D3 . (11)
[D1 , D4 ] b1,3 b1,1 −1
D5
The polynomial vector fields D6 , D8 and D10 are determined by the relations
1 8
D6 = (2[D2 , D4 ] + b2,3 D1 − b2,1 D3 ) − (λ6 D0 − λ4 D2 ) , (12)
4  7 
1 1 3
D8 = 2[D2 , D6 ] + D1 (p3,3 ) + b2,5 D1 − b2,3 D3 − b2,1 D5 − (λ8 D0 − λ4 D4 ) ,
8 2 7
1 4
D10 = (2[D2 , D8 ] + D1 (p3,5 )D1 − b2,5 D3 − b2,3 D5 ) − (λ10 D0 − λ4 D6 ) .
12 21
The commutation relations are obtained from the relations in Theorem 7.1 by the cor-
respondence
L k = ϕ∗ D k , ϕ∗ pi,j = 2℘i,j ϕ∗ qi,j,k = 2℘i,j,k .
Proof. This Theorem follows from Section 9 in [5]. Note that the relations (11) deter-
mine the polynomials Dk (bi,j ) for k ∈ {2, 4}, i ∈ {2, 3}, j ∈ {1, 3, 5}. For Lk given in
Theorem 7.1 the relation (7) holds. Cf. eq. (5), (10), and (12). 
The polynomial dynamical systems in C9 determined by the vector fields Dk are pre-
sented in Section 8 of [13].
9
 8. Lie algebra of derivations of hyperelliptic functions: genus 4 case
Example 8.1. In the case of genus g = 4 Lemma 2.1 implies
   
[L2 , L4 ] 24λ6 −24λ4 0 18 0 0 0 0
 [L2 , L6 ]   20λ8 0 −20λ4 0 36 0 0 0
  1 
 [L2 , L8 ]  16λ10 0 0 −16λ4 0 54 0 0e
  =   L, (13)
[L2 , L10 ] 9 12λ12 0 0 0 −12λ4 0 72 0
[L , L ]  8λ 0 0 0 0 −8λ4 0 90
2 12 14
[L2 , L14 ] 4λ10 0 0 0 0 0 −4λ4 0

⊤
where L e = L0 L2 L4 L6 L8 L10 L12 L14 .

We denote the matrix (c2k e

2,2j (λ)) by C2 (λ). The right hand side of (13) is C2 (λ)L.

Theorem 8.2 (cf. Corollary 6.2 in [15]). In the case of genus g = 4 the following
vector fields give a solution to Problem 1.1
L1 = ∂1 , L3 = ∂3 , L5 = ∂5 , L7 = ∂7 ,
L0 = L0 − t1 ∂1 − 3t3 ∂3 − 5t5 ∂5 − 7t7 ∂7 ,
   
4 8 4
L2 = L2 − ζ1 ∂1 + λ4 t3 ∂1 − t1 − λ4 t5 ∂3 − 3t3 − λ4 t7 ∂5 − 5t5 ∂7 ,
3 9 9
L4 = L4 − ζ3 ∂1 − ζ1 ∂3 +
   
4 2
+ 2λ6 t3 ∂1 − λ4 t3 − λ6 t5 ∂3 − t1 + 3λ4 t5 − λ6 t7 ∂5 − (3t3 + 5λ4 t7 ) ∂7 ,
3 3
L6 = L6 − ζ5 ∂1 − ζ3 ∂3 − ζ1 ∂5 +
 
5 16 8
+ λ8 t3 ∂1 + λ8 t5 ∂3 − λ4 t3 + 2λ6 t5 − λ8 t7 ∂5 − (t1 + 3λ4 t5 + 4λ6 t7 ) ∂7 ,
3 9 9
 
4
L8 = L8 − ζ7 ∂1 − ζ5 ∂3 − ζ3 ∂5 − ζ1 ∂7 + λ10 t3 − λ12 t5 ∂1 +
3
 
20 10
+ λ10 t5 ∂3 − λ8 t5 − λ10 t7 ∂5 − (λ4 t3 + 2λ6 t5 + 3λ8 t7 ) ∂7 ,
9 9
L10 = L10 − ζ7 ∂3 − ζ5 ∂5 − ζ3 ∂7 +
5 4
+ (λ12 t3 − 2λ14 t5 − λ16 t7 )∂1 + λ12 t5 ∂3 + λ12 t7 ∂5 − (λ8 t5 + 2λ10 t7 )∂7 ,
3 3
L12 = L12 − ζ7 ∂5 − ζ5 ∂7 +
   
2 10 14
+ λ14 t3 − 3λ16 t5 − 2λ18 t7 ∂1 + λ14 t5 − λ16 t7 ∂3 + λ14 t7 ∂5 − λ12 t7 ∂7 ,
3 9 9
   
1 5 7
L14 = L14 − ζ7 ∂7 + λ16 t3 − 4λ18 t5 ∂1 + λ16 t5 − 2λ18 t7 ∂3 + λ16 t7 ∂5 .
3 9 9
We denote this Lie algebra by L. The commutation relations are given in Lemma 8.3,
Lemma 8.4, Lemma 8.5, and Corollary 10.5.
Lemma 8.3. For the commutators in the Lie algebra L we have the relations:
[L0 , Lk ] = kLk , k = 0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14;
[Lk , Lm ] = 0, k, m = 1, 3, 5, 7.
Proof. We use the explicit expressions for Lk and the fact that L0 is the Euler vector
field, thus L0 ζk = kζk . 
10
 Lemma 8.4. For the commutators in the Lie algebra L we have the relations:
   
[L1 , L2 ] ℘1,1 −1 0 0
 [L1 , L4 ]  ℘1,3 ℘1,1 −1 0  
    L
 [L1 , L6 ]  ℘1,5 ℘1,3 ℘1,1 −1   1 
    L3 
 [L1 , L8 ]  = ℘1,7 ℘1,5 ℘1,3 ℘1,1   L5  ,
[L , L ]  0 ℘ 
 1 10   1,7 ℘1,5 ℘1,3 
[L , L ]  0 L7
1 12 0 ℘1,7 ℘1,5 
[L1 , L14 ] 0 0 0 ℘1,7
     
[L3 , L2 ] ℘1,3 − λ4 0 −3 0 7λ4
 [L3 , L4 ]   ℘3,3 ℘1,3 − λ4 0 −3     6λ6 
    L1  
 [L3 , L6 ]   ℘3,5 ℘3,3 ℘1,3 − λ4 0    1  5λ8 
    L3   
 [L3 , L8 ]  =  ℘3,7 ℘3,5 ℘3,3 ℘1,3 − λ4    + 4λ10  L1 ,
[L , L ]  L
 3 10   0 ℘3,7 ℘3,5 ℘3,3   L
5 3 3λ 
 12 
[L , L ]  0 0 ℘3,7 ℘3,5  7 2λ 
3 12 14
[L3 , L14 ] 0 0 0 ℘3,7 λ16
     
[L5 , L2 ] ℘1,5 0 0 −5 2λ4
 [L5 , L4 ]   ℘3,5 ℘1,5 −3λ4 0    3λ6 
    L1  
 [L5 , L6 ]   ℘5,5 ℘3,5 ℘1,5 − 2λ6 −3λ4   
4  4λ8 
    L  
 [L5 , L8 ]  = ℘5,7 − λ12 ℘5,5 ℘3,5 − λ8 ℘1,5 − 2λ6   3
L5  + 9  5λ10  L3 ,
[L , L ]  −2λ ℘5,7 − λ12 ℘5,5 
℘3,5 − λ8   
 5 10   14
L 6λ12 
[L , L ]  −3λ −2λ14 ℘5,7 ℘5,5  7 7λ 
5 12 16 14
[L5 , L14 ] −4λ18 −3λ16 0 ℘5,7 8λ16
     
[L7 , L2 ] ℘1,7 0 0 0 2λ4
 [L7 , L4 ]   ℘3,7 ℘1,7 0 −5λ4     3λ6 
    L  
 [L7 , L6 ]   ℘5,7 ℘3,7 ℘1,7 −4λ6   1   4λ 
    L3  2  8 
 [L7 , L8 ]  =  ℘7,7 ℘5,7 ℘3,7 ℘1,7 − 3λ8    + 5λ10  L5 .
[L , L ]  −λ L5
 7 10   16 ℘7,7 ℘5,7 ℘3,7 − 2λ10 
 L
9 6λ 
 12 
[L , L ] −2λ −λ16 ℘7,7 ℘5,7 − λ12  7 7λ 
7 12 18 14
[L7 , L14 ] 0 −2λ18 −λ16 ℘7,7 8λ16

The proof follows from the explicit expressions for Lk .

Lemma 8.5. For the commutators in the Lie algebra L we have the relations
 
L0
   
[L2 , L4 ]  L2  −℘1,1,3 ℘1,1,1 0 0  
 
 [L2 , L6 ]   L4   −℘1,3,3 − ℘1,1,5 ℘1,1,3 ℘1,1,1 0  L1
    1 
 [L2 , L8 ]   L6  −2℘1,3,5 − ℘1,1,7 ℘1,1,5 ℘1,1,3 ℘1,1,1  L3 
  = C2 (λ)   +   ,
[L2 , L10 ]  L8  2 −2℘1,3,7 − ℘1,5,5 ℘1,1,7 ℘1,1,5 ℘1,1,3  L5 

[L , L ] L   −2℘1,5,7 0 ℘1,1,7 ℘1,1,5  L7
2 12  10 
[L2 , L14 ] L12  −℘1,7,7 0 0 ℘1,1,7
L14

where the polynomial matrix C2 (λ) = (c2k

2,2j (λ)) is given in (13).

Proof. From Theorem 6.3 in [15] we obtain the expressions for L2k ζs , where s = 1, 3, 5, 7
and k = 1, 2, 3, 4, 5, 6, 7. Provided this, the proof follows from the explicit expressions
for Lk . 
11
 9. Generators in the polynomial Lie algebra in C12
In the case of genus g = 4 the expressions for the coordinates (λ) and (p) in bi,j obtained
in Corollaries 3.4 and 3.5 are the following:
1 1 1 1
λ4 = −3b21,1 + b3,1 − 2b1,3 , λ6 = 2b31,1 + b22,1 − b1,1 b3,1 − 2b1,1 b1,3 + b3,3 − 2b1,5 ,
2 4 2 2
2 1 1
λ8 = 4b1,1 b1,3 + b1,3 b1,3 − 2b1,1 b1,5 + b3,5 − 2b1,7 − (b3,1 b1,3 − b2,1 b2,3 + b1,1 b3,3 ), (14)
2 2
1 1
λ10 = 2(b1,1 b21,3 + b1,5 b1,3 − b1,1 b1,7 + 2b21,1 b1,5 ) + b22,3 − (b1,3 b3,3 − b3,7 + b3,1 b1,5 − b2,1 b2,5 + b1,1 b3,5 ),
4 2
2
λ12 = 4b1,1 (b1,3 b1,5 + b1,1 b1,7 ) + b1,5 + 2b1,3 b1,7 −
1
− (b3,3 b1,5 − b2,3 b2,5 + b1,3 b3,5 + b3,1 b1,7 − b2,1 b2,7 + b1,1 b3,7 ) ,
2
1 1 1
λ14 = 2b1,1 b21,5 + b22,5 − b1,5 b3,5 + 4b1,1 b1,3 b1,7 + 2b1,5 b1,7 − (b3,3 b1,7 − b2,3 b2,7 + b1,3 b3,7 ) ,
4 2 2
1 1 1
λ16 = 4b1,1 b1,5 b1,7 + b21,7 − (b3,5 b1,7 − b2,5 b2,7 + b1,5 b3,7 ) , λ18 = 2b1,1 b21,7 + b22,7 − b1,7 b3,7 ,
2 4 2
p3,3 = 6b1,1 b1,3 − b3,3 + 6b1,5 , p3,5 = 6b1,1 b1,5 − b3,5 + 6b1,7 , p3,7 = 6b1,1 b1,7 − b3,7 ,


p5,5 = b3,1 b1,5 − b2,1 b2,5 + b1,1 b3,5 − 2 4b21,1 + b1,3 b1,5 + 10b1,1 b1,7 − 2b3,7 ,


p5,7 = b3,1 b1,7 − b2,1 b2,7 + b1,1 b3,7 − 2 4b21,1 + b1,3 b1,7 ,
p7,7 = b3,3 b1,7 − b2,3 b2,7 + b1,3 b3,7 − 2 (4b1,1 b1,3 + b1,5 ) b1,7 .
Theorem 9.1. In the case of genus g = 4 the homogeneous polynomial vector fields Ds
for s ∈ {0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14} give a solution to Problem 1.2. The vector fields
D1 , D3 , D5 , D7 , and D0 are determined by Lemmas 4.2, 4.3 and 4.6 for g = 4. We
set p′i,j = D1 (pi,j ) and p′′i,j = D1 (D1 (pi,j )) for i, j ∈ {3, 5, 7}. The polynomial vector
fields D2 and D4 are determined by the conditions
1 7
D2 (b1,1 ) = −b21,1 + b3,1 + 2b1,3 + λ4 , (15)
2 9
1 4 1
D2 (b1,3 ) = −b1,1 b1,3 + b3,3 + 3b1,5 − λ4 b1,1 + p3,3 ,
2 3 2
1 8 1
D2 (b1,5 ) = −b1,1 b1,5 + b3,5 + 5b1,7 − λ4 b1,3 + p3,5 ,
2 9 2
1 4 1
D2 (b1,7 ) = −b1,1 b1,7 + b3,7 − λ4 b1,5 + p3,7 ,
2 9 2
2
D4 (b1,1 ) = −2b1,1 b1,3 + b3,3 + 2b1,5 + λ6 ,
3
1 1 1
D4 (b1,3 ) = −b21,3 + 3b1,7 + λ4 b1,3 − 2λ6 b1,1 − λ8 − b1,1 p3,3 + p′′3,3 + p3,5 ,
2 2 2
4 1 1 ′′ 1
D4 (b1,5 ) = −b1,3 b1,5 + 3λ4 b1,5 − λ6 b1,3 − b1,1 p3,5 + p3,5 + p5,5 ,
3 2 2 2
2 1 1 ′′ 1
D4 (b1,7 ) = −b1,3 b1,7 + 5λ4 b1,7 − λ6 b1,5 − b1,1 p3,7 + p3,7 + p5,7 ,
3 2 2 2
and the relations  
    D1
[D1 , D2 ] b −1 0  
= 1,1 D3 . (16)
[D1 , D4 ] b1,3 b1,1 −1
D5
12
 The polynomial vector fields D6 , D8 and D10 are determined by the relations
1 4
D6 = (2[D2 , D4 ] + b2,3 D1 − b2,1 D3 ) − (λ6 D0 − λ4 D2 ) ,
4   3 
1 1 ′ 5
D8 = 2[D2 , D6 ] + p3,3 + b2,5 D1 − b2,3 D3 − b2,1 D5 − (λ8 D0 − λ4 D4 ) , (17)
8 2 9
1

8
D10 = 2[D2 , D8 ] + p′3,5 + b2,7 D1 − b2,5 D3 − b2,3 D5 − b2,1 D7 − (λ10 D0 − λ4 D6 ) ,
12    27

1 ′ 1 ′ 1
D12 = 2[D2 , D10 ] + p3,7 + p5,5 D1 − b2,7 D3 − b2,5 D5 − b2,3 D7 − (λ12 D0 − λ4 D8 ) ,
16 2 6
1
4
D14 = 2[D2 , D12 ] + p′5,7 D1 − b2,7 D5 − b2,5 D7 − (λ14 D0 − λ4 D10 ) .
20 45
We denote by D the polynomial Lie algebra generated by the polynomial vector fields
Ds for s ∈ {0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14}. The commutation relations for D are given
in Lemma 10.1, Lemma 10.2, Lemma 10.3, and Lemma 10.4.
Proof. First we note that the polynomial vector fields Ds are determined uniquely by
the Theorem. The polynomial vector fields D1 , D3 , D5 , D7 , and D0 are determined by
Lemmas 4.2, 4.3 and 4.6. Relations (15) determine the polynomials D2 (b1,j ) and D4 (b1,j )
for j ∈ {1, 3, 5, 7}. Relations (16) determine the polynomials D2 (bi,j ) and D4 (bi,j ) for
i ∈ {2, 3}, j ∈ {1, 3, 5, 7}. Relations (17) determine the polynomial vector fields D2k
for k ∈ {3, 4, 5, 6, 7} given the polynomial vector fields Ds for s ∈ {1, 3, 5, 7} and D2m
for m < k.
Thus we have a system of 12 polynomial vector fields that are explicitly defined in
the coordinates (b). By Corollary 4.4, the polynomial vector fields Ds for s ∈ {1, 3, 5, 7}
are projectable for the polynomial map ρ : C12 → C8 . Their pushforwards are zero. By
Corollary 4.7, the polynomial vector field D0 is projectable for the polynomial map ρ
with pushforward L0 . It is a direct calculation to check that the polynomial vector fields
D2k for k ∈ {1, 2, 3, 4, 5, 6, 7} are projectable for the polynomial map ρ with pushfor-
wards L2k . This proves the claim of the Theorem that the vector fields Ds give a solution
to Problem 1.2.
The commutation relations given in Section 10 in Lemmas 10.1, 10.2, 10.3, and 10.4
are proved by direct computation. 
Theorem 9.2. For the polynomial vector fields Dk for k ∈ {0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14},
determined by Theorem 9.1, we have
Lk (ϕ∗ bi,j ) = ϕ∗ Dk (bi,j ), (18)
where the vector fields Lk are given in Theorem 8.2.
Proof. For the polynomial vector fields D1 , D3 , D5 , D7 , and D0 the claim of the Theorem is
given by Lemmas 4.2, 4.3, and 4.6. For k ∈ {1, 2, 3, 4, 5, 6, 7} define the vector fields D e2s
by the relation (18). The vector fields D e2s are linear combinations of the fields Dk
for k ∈ {0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14} with coefficients rational functions in (b). The
pushforward of D e2s is L2s , which is equal to the pushforward of D2s . Therefore we have
e2s = D2s + α1 (b)D1 + α3 (b)D3 + α5 (b)D5 + α7 (b)D7
D 2s 2s 2s 2s
j
for some rational functions α2s (b).
Comparing results of Lemma 8.4 and Lemma 10.2,
j j
we obtain Di (α2s (b)) = 0 for i ∈ {1, 3, 5, 7}, thus α2s (b) is a rational function in (λ).
j
We have wt α2s (b) = 2s − j, where j is odd, while wt λm is even for all m (see (14)), thus
j
we obtain α2s e2s = D2s .
(b) = 0 and D 
13
 10. Commutation relations in the polynomial Lie algebra in C12
Lemma 10.1. The relations hold for the commutators in the polynomial Lie algebra D:
[D0 , Dk ] = kDk , k = 0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14;
[Dk , Dm ] = 0, k, m = 1, 3, 5, 7.
Lemma 10.2. The relations hold for the commutators in the polynomial Lie algebra D:
   
[D1 , D2 ] b1,1 −1 0 0
 [D1 , D4 ]  b1,3 b1,1 −1 0   
    D
 [D1 , D6 ]  b1,5 b1,3 b1,1 −1   1 
    D3 
 [D1 , D8 ]  = b1,7 b1,5 b1,3 b1,1   , (19)
    D5 
[D1 , D10 ]  0 b1,7 b1,5 b1,3  D
[D , D ]  0 0 b1,7 b1,5  7
1 12
[D1 , D14 ] 0 0 0 b1,7
     
[D3 , D2 ] 2b1,3 − 2λ4 0 −6 0 7λ4
 [D3 , D4 ]   p3,3 2b1,3 − 2λ4 0 −6    6λ6 
    D1  
 [D3 , D6 ]  1  p3,5 p3,3 2b1,3 − 2λ4 0    1  5λ8 
    D3   
 [D3 , D8 ]  =  p3,7 p3,5 p3,3 2b1,3 − 2λ4  
  + 4λ10  D1 ,
  2  D 5 3 3λ 
[D3 , D10 ]  0 p3,7 p3,5 p3,3  D  12 
[D , D ]  0 0 p3,7 p3,5  7 2λ 
3 12 14
[D3 , D14 ] 0 0 0 p3,7 λ16
     
[D5 , D2 ] 2b1,5 0 0 −10 2λ4
 [D5 , D4 ]   p3,5 2b1,5 −6λ4 0    3λ6 
    D1  
 [D5 , D6 ]  1  p5,5 p3,5 2b1,5 − 4λ6 −6λ4   
4  4λ8 
    D  
 [D5 , D8 ]  = p5,7 − 2λ12 p5,5 p3,5 − 2λ8 2b1,5 − 4λ6   3
+ 5λ10  D3 ,
  2   D5  9  
[D5 , D10 ]  −4λ14 p5,7 − 2λ12 p5,5 p3,5 − 2λ8 
D 6λ12 
[D , D ]  −6λ −4λ14 p5,7 p5,5  7 7λ 
5 12 16 14
[D5 , D14 ] −8λ18 −6λ16 0 p5,7 8λ16
     
[D7 , D2 ] 2b1,7 0 0 0 2λ4
 [D7 , D4 ]   p3,7 2b1,7 0 −10λ4     3λ6 
    D  
 [D7 , D6 ]  1  p5,7 p3,7 2b1,7 −8λ6   1   4λ 
    D3  2  8 
 [D7 , D8 ]  =  p7,7 p5,7 p3,7 2b1,7 − 6λ8    + 5λ10  D5 .
  2  D5 9 6λ 
[D7 , D10 ] −2λ16 p7,7 p5,7 p3,7 − 4λ10 
D7  12 
[D , D ] −4λ  7λ 
7 12 18 −2λ16 p7,7 p5,7 − 2λ12 14
[D7 , D14 ] 0 −4λ18 −2λ16 p7,7 8λ16
Lemma 10.3. The relations hold for the commutators in the polynomial Lie algebra D:
 
D0  
 
[D2 , D4 ]  D2  −b2,3 b2,1 0 0  
   − 1 p′3,3 − b2,5 b2,3 b2,1 0  D1
 [D2 , D6 ]   D4 
    1 2 
 −p′3,5 − b2,7 b2,5 b2,3 b2,1  D3 
 [D2 , D8 ]   D6 
  = C2 (λ)  +  ′  ,
[D2 , D10 ]  D8  2  −p3,7 − 12 p′5,5 b2,7 b2,5 b2,3   
 D5
[D , D ] D   −p5,7′
0 b2,7 b2,5  D7
2 12  10 
[D2 , D14 ] D  − 1 ′
p 0 0 b
12 2 7,7 2,7
D14
where the polynomial matrix C2 (λ) = (c2k
2,2j (λ)) is given in (13).

14
 Lemma 10.4. Set p1,k = 2b1,k and qi,j,k = Di (pj,k ) for i, j, k ∈ {1, 3, 5, 7}. The relations
for m, l ∈ 2, 3, 4, 5, 6, m 6 l hold for the commutators in the polynomial Lie algebra D:
7
X
⊤
[D2m , D2l+2 ] = c2k
2m,2l+2 (λ)D2k + (A2m,2l+2 ) D1 D3 D5 D7 ,
k=0
where
   
A4,6 −q3,3,3 q1,3,3 − 2q1,1,5 2q1,1,3 0
 A4,8   −2q3,3,5 −2q1,1,7 2q1,3,3 2q1,1,3 
   
 A4,10  −2q3,3,7 − q3,5,5 −q1,5,5 2q1,3,5 2q1,3,3 
   
 A4,12   −2q3,5,7 −2q1,5,7 2q1,3,7 2q1,3,5 
   
 A4,14   −q3,7,7 −q1,7,7 0 2q1,3,7 
   
 A6,8   q3,3,7 − 2q3,5,5 2q1,5,5 − q3,3,5 − 2q1,3,7 q3,3,3 − 2q1,1,7 q1,3,3 + 2q1,1,5 
   
 A6,10  1 −q5,5,5 − 2q3,5,7 2q1,5,7 − q3,5,5 − q3,3,7 q3,3,5 + q1,5,5 − 2q1,3,7 q3,3,3 + 2q1,3,5 
   
 A6,12  =  −2q5,5,7 −2q3,5,7 q3,3,7 q3,3,5 + 2q1,5,5  .
  4 
 A6,14   −q5,7,7 −q3,7,7 −q1,7,7 q3,3,7 + 2q1,5,7 
A  −2q 2q3,3,5 − q1,5,5 
 8,10   3,7,7 − q5,5,7 −q5,5,5 + 2q1,7,7 2q1,5,7 + q3,5,5 − 2q3,3,7 
A   −2q −2q 2q1,7,7 2q3,5,5 
 8,12   5,7,7 5,5,7 
A   −q −q −q 2q3,5,7 + q1,7,7 
 8,14   7,7,7 5,7,7 3,7,7 
A   0 −2q 2q − q5,5,7 q5,5,5 
 10,12   5,7,7 3,7,7 
A10,14   0 −q7,7,7 −q5,7,7 q5,5,7 + q3,7,7 
A12,14 0 0 −q7,7,7 q5,7,7
Corollary 10.5 (From Lemma 10.4 and Theorem 9.2). The relations for m, l ∈
2, 3, 4, 5, 6, m 6 l hold for the commutators in the polynomial Lie algebra L:
7
X
⊤
[L2m , L2l+2 ] = c2k
2m,2l+2 (λ)L2k + (A2m,2l+2 ) L1 L3 L5 L7 ,
k=0
where
   
A4,6 −℘3,3,3 ℘1,3,3 − 2℘1,1,5 2℘1,1,3 0
 A4,8   −2℘3,3,5 −2℘1,1,7 2℘1,3,3 2℘1,1,3 
   
 A4,10  −2℘3,3,7 − ℘3,5,5 −℘1,5,5 2℘1,3,5 2℘1,3,3 
   
 A4,12   −2℘3,5,7 −2℘1,5,7 2℘1,3,7 2℘1,3,5 
   
 A4,14   −℘3,7,7 −℘1,7,7 0 2℘1,3,7 
   
 A6,8   ℘3,3,7 − 2℘3,5,5 2℘1,5,5 − ℘3,3,5 − 2℘1,3,7 ℘3,3,3 − 2℘1,1,7 ℘1,3,3 + 2℘1,1,5 
   
 A6,10  1 −℘5,5,5 − 2℘3,5,7 2℘1,5,7 − ℘3,5,5 − ℘3,3,7 ℘3,3,5 + ℘1,5,5 − 2℘1,3,7 ℘3,3,3 + 2℘1,3,5 
   
 A6,12  =  −2℘5,5,7 −2℘3,5,7 ℘3,3,7 ℘3,3,5 + 2℘1,5,5  .
  2 
A
 6,14   −℘ 5,7,7 −℘ 3,7,7 −℘ 1,7,7 ℘3,3,7 + 2℘1,5,7 
A  −2℘ 2℘3,3,5 − ℘1,5,5 
 8,10   3,7,7 − ℘5,5,7 −℘5,5,5 + 2℘1,7,7 2℘1,5,7 + ℘3,5,5 − 2℘3,3,7 
A   −2℘5,7,7 −2℘5,5,7 2℘1,7,7 2℘3,5,5 
 8,12   
A   −℘7,7,7 −℘5,7,7 −℘3,7,7 2℘3,5,7 + ℘1,7,7 
 8,14   
A   0 −2℘ 2℘ 
 10,12   5,7,7 3,7,7 − ℘5,5,7 ℘5,5,5 
A10,14   0 −℘7,7,7 −℘5,7,7 ℘5,5,7 + ℘3,7,7 
A12,14 0 0 −℘7,7,7 ℘5,7,7

15
 11. Polynomial dynamical systems in C12 related to differentiations
of genus 4 hyperelliptic functions
As we have noted throughout the work, the generators Lk of Der F and the polynomial
vector fields Dk in C3g are related by (7). The polynomial vector fields D0 , D1 , and Ds
for s ∈ {3, 5, . . . , 2g − 1} are given is Section 4. Section 5 gives the graded homogeneous
polynomial dynamical systems S0 , S1 , and Ss for s ∈ {3, 5, . . . , 2g − 1} in C3g determined
by these vector fields.
In this Section in the case of genus g = 4 we determine the remaing graded homogeneous
polynomial dynamical systems S2k for k ∈ {1, 2, 3, 4, 5, 6, 7}. By definition, the dynamical
system S2k corresponding to the vector field D2k is given by
∂
bi,j = D2k (bi,j ).
∂τ2k
It has been noted in [2] in the case of genus g = 2 that to determine such systems it is
sufficient to determine the polynomials D2k (b1,1 ). Indeed, we have the relation
D2k (λm ) = L2k (λm )
for m ∈ {4, 6, 8, 10, 12, 14, 16, 18} that determines the action of D2k on the coordinates (λ).
The relations (14) imply
3 1 1
b1,3 = − b21,1 + b3,1 − λ4 ,
2 4 2
1 1 1 1
b1,5 = b31,1 + b22,1 − b1,1 b3,1 − b1,1 b1,3 + b3,3 − λ6 ,
8 4 4 2
1 1 1 1
b1,7 = 2b21,1 b1,3 + b1,3 b1,3 − b1,1 b1,5 + b3,5 − (b3,1 b1,3 − b2,1 b2,3 + b1,1 b3,3 ) − λ8 .
2 4 4 2
Corollary 4.4 implies D1 (λm ) = 0 and Lemma 4.2 implies
b2,s = D1 (b1,s ), b3,s = D1 (D1 (b1,s )), for s ∈ {1, 3, 5, 7}.
(n)
Finally, the relation (19) determines D2k (D1 (b1,1 )) provided D2k (b1,1 ) is given. To con-
clude, we give explicitly the values D2k (b1,1 ) for k ∈ {1, 2, 3, 4, 5, 6, 7}. By (15) we have
1 7
D2 (b1,1 ) = −b21,1 + 2b1,3 + b3,1 + λ4 ,
2 9
2
D4 (b1,1 ) = b3,3 − 2b1,1 b1,3 + 2b1,5 + λ6 .
3
The remaining values are
1 1 3 5
D6 (b1,1 ) = −b21,3 + 2b1,7 − 2b1,1 b1,5 + b3,1 b1,3 − b3,3 b1,1 + b3,5 + λ8 ,
2 2 2 9
4
D8 (b1,1 ) = b1,5 b3,1 − b1,1 b3,5 − 2b1,3 b1,5 − 2b1,1 b1,7 + 2b3,7 + λ10 ,
9
1 1 3 3 1
D10 (b1,1 ) = −b21,5 − 2b1,3 b1,7 + b3,3 b1,5 − b3,5 b1,3 + b3,1 b1,7 − b3,7 b1,1 + λ12 ,
2 2 2 2 3
2
D12 (b1,1 ) = b1,7 b3,3 − b1,3 b3,7 − 2b1,5 b1,7 + λ14 ,
9
1 1 1
D14 (b1,1 ) = −b21,7 + b3,5 b1,7 − b3,7 b1,5 + λ16 .
2 2 9

16
 References
[1] V. M. Buchstaber, Multidimensional Sigma Functions and Applications, Victor Enolski (1945–
2019), Notices Amer. Math. Soc., 67:11 (2020), 1756–1760.
[2] V. M. Buchstaber, D. V. Leikin, Solution of the Problem of Differentiation of Abelian Functions
over Parameters for Families of (n, s)-Curves, Funct. Anal. Appl., 42:4 (2008), 268–278.
[3] B. A. Dubrovin, S. P. Novikov, A periodic problem for the Korteweg-de Vries and Sturm-Liouville
equations. Their connection with algebraic geometry. (Russian) Dokl. Akad. Nauk SSSR, 219:3
(1974), 531–534.
[4] F. G. Frobenius, L. Stickelberger, Über die Differentiation der elliptischen Functionen nach
den Perioden und Invarianten, J. Reine Angew. Math., 92 (1882), 311–337.
[5] E. Yu. Bunkova, Differentiation of genus 3 hyperelliptic functions, European Journal of Mathe-
matics, 4:1 (2018), 93–112.
[6] V. M. Buchstaber, V. Z. Enolski, D. V. Leykin, σ-functions: old and new results, Integrable
Systems and Algebraic Geometry, vol. 2, LMS Lecture Note Series, no. 459, Cambridge Univ. Press,
2020, 175–214.
[7] H. F. Baker, On the hyperelliptic sigma functions, Amer. Journ. Math., 20 (1898), 301–384.
[8] V. M. Buchstaber, V. Z. Enolskii, D. V. Leikin, Kleinian functions, hyperelliptic Jacobians and
applications, Reviews in Mathematics and Math. Physics, 10:2, Gordon and Breach, London, 1997,
3–120.
[9] V. M. Buchstaber, V. Z. Enolskii, D. V. Leikin, Hyperelliptic Kleinian functions and applica-
tions, “Solitons, Geometry and Topology: On the Crossroad”, Adv. Math. Sci., AMS Transl., 179:2,
Providence, RI, 1997, 1–34.
[10] V. M. Buchstaber, V. Z. Enolskii, D. V. Leikin, Multi-Dimensional Sigma-Functions, arXiv:
1208.0990, 2012, 267 pp.
[11] E. T. Whittaker, G. N. Watson, A Course of Modern Analysis, Reprint of 4th (1927) ed., Vol 2.
Transcendental functions, Cambridge Univ. Press, Cambridge, 1996.
[12] V. M. Buchstaber, Polynomial dynamical systems and Korteweg–de Vries equation, Proc. Steklov
Inst. Math., 294 (2016), 176–200.
[13] V. M. Buchstaber, E. Yu. Bunkova, Hyperelliptic Sigma functions and Adler–Moser polynomi-
als, Functional Analysis and Its Applications, 55:3 (2021), 3–25.
[14] V. M. Buchstaber, E. Yu. Bunkova, Differentiation of Genus 4 Hyperelliptic Functions, 2019,
arXiv: 1912.11379
[15] V. M. Buchstaber, E. Yu. Bunkova, Sigma Functions and Lie Algebras of Schrödinger Opera-
tors, Functional Analysis and Its Applications, 54:4 (2020), 229–240.
[16] V. M. Buchstaber, D. V. Leikin, Polynomial Lie Algebras, Funct. Anal. Appl., 36:4 (2002), 267–
280.
[17] V. I. Arnold, Singularities of Caustics and Wave Fronts, Mathematics and its Applications, vol. 62,
Kluwer Academic Publisher Group, Dordrecht, 1990.
[18] V. M. Buchstaber, D. V. Leikin, Heat Equations in a Nonholomic Frame, Funct. Anal. Appl.,
38:2 (2004), 88–101.

Steklov Mathematical Institute of Russian Academy of Sciences, Moscow, Russia

Email address: buchstab@mi-ras.ru, bunkova@mi-ras.ru

17