The notion of “Siegel domains” was introduced by [8]. It was then shown that every homogeneous bounded domain is holomorphically equivalent to a Siegel domain (of the second kind) determined uniquely up to an affine isomorphism ([15], cf. also [2], [4], [9b]). In a recent note [10b], I have shown that among (homogeneous) Siegel domains the symmetric domains can be characterized by three conditions (i), (ii), (iii) on the data (U, V, Ω, F) defining the Siegel domain (see Theorem in § 2 of this paper). The class of homogeneous Siegel domains satisfying partial conditions (i), (ii), which we propose to call “quasi-symmetric”, seems to be of some interest, since for instance the fibers appearing in the expressions of symmetric domains as Siegel domains of the third kind fall in this class ([10b], [16]).

Let K be a Galois extension of an algebraic number field k of finite degree with Galois group g. Then g acts on a congruent ideal class group of K as a group of automorphisms, when the class field M over K corresponding to is normal over K. Let Ig be the augmentation ideal of the group ring Zg over the ring of integers Z, namely Ig be the ideal of Zg generated by σ − 1, σ running over all elements of g. Then is the group of all elements aσ-1 where a and σ belong to and g respectively.

Let N be a positive integer and let Γ0(N) be the subgroup of SL(2, Z) defined by all matrices with c ≡ 0 (mod N). Let S2(Γ0(N)) be the space of holomorphic cusp forms of weight 2 with respect to Γ0(N) and let be the “essential part” of S2(Γ0(N)), which is defined in [1].

Krull’s principal ideal theorm [Krull] states that q elements in the maximal ideal of a local noetherian ring generate an ideal whose minimal components are all of height at most q. Writing R for the ring, we may consider the q elements, x1, · · ·, xq say, as coordinates of an element x ∈ Rq.

Let S (resp. S′) be a (real) hypersurface (i.e. a real analytic sub-manifold of codimension 1) of an n-dimensional complex manifold M (resp. M′). A homeomorphism f of S onto S′ is called a pseudo-conformal homeomorphism if it can be extended to a holomorphic homeomorphism of a neighborhood of S in M onto a neighborhood of S′ in M. In case such an f exists, we say that S and S′ are pseudo-conformally equivalent. A hypersurface S is called non-degenerate (index r) if its Levi-form is non-degenerate (and its index is equal to r) at each point of S.

Let S be a compact Riemann surface and let Sn be the surface obtained from S in the course of a pinching deformation. We denote by Γn the quasi-Fuchsian group representing Sn in the Teichmüller space T(Γ), where Γ is a Fuchsian group with U/Γ = S (U: the upper half plane). Then in the previous paper [7] we showed that the limit of the sequence of Γn is a cusp on the boundary ∂T(Γ). In this paper we will consider the case of Schottky space . Let Gn be a Schottky group with Ω(Gn)/Gn = Sn. Then the purpose of this paper is to show what the limit of Gn is.

Dans les travaux célèbres [1], [2], [3], E. Cartan a fait des recherches sur les pseudo-groupes, sans distinguer entre les deux cas, transitif ou intransitif. Tout d’abord, Cartan a abouti au théorème qu’il a considéré comme la base de la théorie des pseudo-groupes continus, c’est-à-dire le premier théorème fondamental: Tout pseudo-groupe de Lie admet un prolongement holoédrique opérant sur un certain nombre r de variables xi et défini comme l’ensemble des transformations qui laissent invariants

1° un certain nombre de fonctions des xi;

2° r formes de Pfaff wi(x, y, dx) linéairement indépendantes par rapport aux différentielles dxi et dont les coefficients peuvent dépendre d’autres variables auxiliaires yj; enfin le prolongement considéré a ses équations de définition du premier ordre (voir [5]). Puis, il a introduit l’équation structurale de manière à caractériser le pseudo-groupe.