The term “simplest” field has been used to describe certain totally real, cyclic number fields of degrees 2, 3, 4, 5, 6, and 8. For each of these degrees, the fields are defined by a one-parameter family of polynomials with constant term ±1. The regulator of these “simplest” fields is small in an asymptotic sense: in consequence, the class number of these fields tends to be large.

In this paper we will study some aspects of non-commutative invariant theory. Let V be a finite-dimensional vector space over a field K of characteristic zero and let

K[V] = K⊕V⊕S2(V)⊕…, and

K′V› = K⊕V⊕⊕2(V)⊕⊕3V⊕&

be respectively the symmetric algebra and the tensor algebra over V. Let G be a subgroup of GL(V). Then G acts on K[V] and K′V›. Much of this paper is devoted to the study of the (non-commutative) invariant ring K′V›G of G acting on K′V›.

In the first part of this paper, we shall study the invariant ring in the following situation.

The two main problems in the theory of the theta correspondence or lifting (between automorphic forms on some adelic orthogonal group and on some adelic symplectic or metaplectic group) are the characterization of kernel and image of this correspondence. Both problems tend to be particularly difficult if the two groups are approximately the same size.

Let p0 be a prime, p0 > 3 and Γ0(p0), Γ1(p0) as usual, the congruence subgroups of Γ = PSL2(ℤ).

In [N-Pi-Po] the notion of affine Kähler immersion for complex manifolds has been introduced: if Mn is an n-dimensional complex manifold and f: Mn -→ Cn+1 is a holomorphic immersion together with an anti-holomorphic transversal vector field ζ, we can induce a connection ▽ on Mn which is Kähler-like, that is, its curvature tensor R satisfies R(Z, W) = 0 as long as Z, W are (1, 0) complex vector fields on M.

Let h1(u),…, hk(u) be Laurent polynomials of m-variables and let

be a non-degenerate complete intersection variety. Such an intersection variety appears as an exceptional divisor of a resolution of non-degenerate complete intersection varieties with an isolated singularity at the origin (Ok4]).

Let K be a field of characteristic zero and let

be m generic n by n matrices over K. That is, xij(k) are independent commuting indeterminates over K. The K-subalgebra generated by X1…, Xm is called a ring of n by n generic matrices and is denoted by R(n, m).

In this paper we shall prove the following theorem conjectured by Miyake in [3] (see also Jaulent [2]).

THEOREM. Let k be a finite algebraic number field and K be an unramified abelian extension of k, then all ideals belonging to at least [K: k] ideal classes of k become principal in K.

Since the capitulation homomorphism is equivalently translated to a group-transfer of the galois group (see Miyake [3]), it is enough to prove the following group-theoretical verison:

In [5], H. Maass showed that the dimension of a space of generalized Whittaker functions satisfying certain system of differential equations on Siegel’s upper half space H2 of degree 2 is three. First of all, we shall investigate the structure of a space of generalized Whittaker functions which are eigen functions for the algebra of invariant differential operators on H2. The theory of generalized Whittaker functions is discussed in Yamashita [12], [13], [14], [15] with full generality. But, we will get an outlook of the space of generalized Whittaker functions by using elementary calculus instead of representation theory of Lie groups.

In a recent paper [PS 89], two of the present authors have found a characterization of a certain space (ℐ)* of generalized functionals of white noise, i.e. generalized functionals on ℐ′(R) equipped with the σ-algebra ℬ generated by its cyclinder sets and with the white noise measure μ given by

A measure μ(dx) on R is said to be unimodal with mode a if μ(dx) = cδa(dx) + f(x) dx, where c ≧ 0, δa(dx) is the delta measure at a and f(x) is non-decreasing for x < a and non-increasing for x > a. A measure is said to be unimodal with mode a if pn is non-decreasing for n ≦ a and non-increasing for n ≧ a. A probability measure μ(dx) on R (resp. on Z) is said to be strongly unimodal on R (resp. on Z) if, for every unimodal probability measure η(dx) on R (resp. on Z), the convolution μ*η(dx) is unimodal on R (resp. on Z).