On holomorphic functions of uniformly bounded type on tensor products (Q1879075)

From MaRDI portal
scientific article
Language Label Description Also known as
English
On holomorphic functions of uniformly bounded type on tensor products
scientific article

    Statements

    On holomorphic functions of uniformly bounded type on tensor products (English)
    0 references
    0 references
    22 September 2004
    0 references
    Given locally convex spaces \(E\) and \(F\), \({\mathcal H}(E,F)\) denotes the space of all holomorphic functions from \(E\) into \(F\). Certain subspaces of \({\mathcal H}(E,F)\) arise naturally. Among these are \({\mathcal H}_b(E,F)\), the space of \(f\) in \({\mathcal H}(E,F)\) which are bounded on bounded subsets of \(E\), and \({\mathcal H}_{ub}(E,F)\), the space of all \(f\) in \({\mathcal H}(E,F)\) which are bounded on all multiples of some neighbourhood of \(0\). In general, \({\mathcal H}_{ub}(E,F)\subseteq {\mathcal H}_b(E,F)\subseteq {\mathcal H}(E,F)\). This paper is concerned with determining conditions on Fréchet or DF spaces \(E\), \(F\) and \(G\) that will ensure \({\mathcal H}_b(G\hat\otimes_\pi E,F)={\mathcal H}_{ub}(G\hat\otimes_\pi E,F)\). These conditions are in terms of the linear topological invariants \((\widetilde\Omega)\), \((\overline{\overline{\Omega}})\), \((LB^\infty)\), \((DN)\), \((\overline{DN})\) and \((\underline{DN})\) of \textit{D. Vogt} [J. Reine Angew. Math. 345, 182--200 (1983; Zbl 0514.46003), Manuscr. Math. 37, 269--301 (1982; Zbl 0512.46003)]. The author shows that (a) given a Banach space \(G\), a Fréchet space \(F\) and a quasi-normable Fréchet space \(E\) with an absolute Schauder basis, then \({\mathcal H}_b(G\hat\otimes_\pi E,F) ={\mathcal H}_{ub}(G\hat\otimes_\pi E,F)\) if (i) \(E\) has \((\widetilde\Omega)\) and \(F\) has \((DN)\), or (ii) \(E\) has \((\overline{\overline{\Omega}})\) and \(F\) has \((\underline{DN})\), (b) given a Fréchet space \(E\), a Fréchet-Montel space \(F\) such that \(F'\) has an absolute basis and a Fréchet space \(G\) with \((\widetilde\Omega)\), then \({\mathcal H}_b(G\hat\otimes_\pi F',E') ={\mathcal H}_{ub}(G\hat\otimes_\pi F',E')\) if (i) \(E\) has \((LB^\infty)\) and \(F\) has \((\overline{DN})\), (ii) \(E\) has \((\widetilde\Omega)\) and \(F\) has \((DN)\), or (iii) \(E\) has \((\overline{\overline{\Omega}})\) and \(F\) has \((\underline{DN})\). (c) if \(E\) is a Fréchet space and \(F\) is a Fréchet-Montel space, then \({\mathcal H}_b(F',E')={\mathcal H}_{ub}(F',E')\) if \(E\) has \((\widetilde\Omega)\) and \(F\) has \((DN)\). This paper contains a number of serious misprints and errors. The first is that the definition of \((\widetilde\Omega)\) given by the author is incorrect. Secondly, in Lemma~4.2 it is not entirely clear which assumptions the author has made about the spaces \(E\) and \(F\). Finally, the author assumes that each point \(z\) in \(\overline{\text{conv}}(U\otimes V_\beta)\), the closed convex hull of \(U\) and \(V_\beta\), has a representation of the form \(z=\sum_{j=1}^\infty\lambda_j x_j\otimes y_j\) with \((x_j)_j\) in \(U\), \((y_j)_j\) in \(V_\beta\) and \(\sum_{j=1}^\infty| \lambda_j| \leq 1\). Such representations of \(z\) are not always possible. This error is not fatal as the proofs of Theorem~1.1 and Theorem~2.1 can easily be adjusted to allow for this fact.
    0 references
    holomorphic functions of uniformly bounded type
    0 references
    linear topological invariants
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references