[go: up one dir, main page]

23 Weierstrass Elliptic and Modular FunctionsModular Functions

§23.15 Definitions

Contents
  1. §23.15(i) General Modular Functions
  2. §23.15(ii) Functions λ⁡(τ), J⁡(τ), η⁡(τ)

§23.15(i) General Modular Functions

In §§23.15–23.19, k and k′ (∈ℂ) denote the Jacobi modulus and complementary modulus, respectively, and q=ei⁢π⁢τ (ℑ⁡τ>0) denotes the nome; compare §§20.1 and 22.1. Thus

23.15.2 k =θ22⁡(0,q)θ32⁡(0,q),
k′ =θ42⁡(0,q)θ32⁡(0,q).

Also 𝒜 denotes a bilinear transformation on τ, given by

23.15.3 𝒜τ=a⁢τ+bc⁢τ+d,

in which a,b,c,d are integers, with

23.15.4 a⁢d−b⁢c=1.

The set of all bilinear transformations of this form is denoted by SL(2,ℤ) (Serre (1973, p. 77)).

A modular function f⁡(τ) is a function of τ that is meromorphic in the half-plane ℑ⁡τ>0, and has the property that for all 𝒜∈SL⁢(2,ℤ), or for all 𝒜 belonging to a subgroup of SL(2,ℤ),

23.15.5 f⁡(𝒜τ)=c𝒜⁢(c⁢τ+d)ℓ⁢f⁡(τ),
ℑ⁡τ>0,

where c𝒜 is a constant depending only on 𝒜, and ℓ (the level) is an integer or half an odd integer. (Some references refer to 2⁢ℓ as the level). If, as a function of q, f⁡(τ) is analytic at q=0, then f⁡(τ) is called a modular form. If, in addition, f⁡(τ)→0 as q→0, then f⁡(τ) is called a cusp form.

§23.15(ii) Functions λ⁡(τ), J⁡(τ), η⁡(τ)

Elliptic Modular Function

23.15.6 λ⁡(τ)=θ24⁡(0,q)θ34⁡(0,q);

compare also (23.15.2).

Klein’s Complete Invariant

23.15.7 J⁡(τ)=(θ28⁡(0,q)+θ38⁡(0,q)+θ48⁡(0,q))354⁢(θ1′⁡(0,q))8,

where (as in §20.2(i))

23.15.8 θ1′⁡(0,q)=∂θ1⁡(z,q)/∂z|z=0.

Dedekind’s Eta Function (or Dedekind Modular Function)

In (23.15.9) the branch of the cube root is chosen to agree with the second equality; in particular, when τ lies on the positive imaginary axis the cube root is real and positive. See also 27.14.12.