On a new characterization of the classical orthogonal polynomials (Q1198940)

The truncated and reversed measure \(\mu_ n^ R\) is defined on \([0;1]\), \([0;\infty)\), or \((-\infty;\infty)\) as a measure corresponding to \((2n- 1)\)-th approximant of a \(g\)-fraction, \(n\)-th approximant of a modified \(S\)-fraction, or \(n\)-th approximant of a \(J\)-fraction correspondingly [see \textit{W. B. Jones} and \textit{W.J. Thron}, Continued fractions: Analytic theory and applications (1980; Zbl 0445.30003)] with the reversed sequence of coefficients. The main result asserts that this measure has equal masses on its \(n\) support points for all \(n\in\mathbb{N}\) if and only if the origin measure has Jacobi, Laguerre, or Hermite density proportional to \(x^ \beta(1-x)^ \alpha\) (\(\alpha,\beta>-1\)), \((1/\beta)(x/\beta)^ \alpha e^{-x/\beta}\) (\(\alpha>-1\), \(\beta>0\)), or \(\exp[-((x-b)/\sqrt{a})^ 2]\) (\(a>0\), \(b\in\mathbb{R}\)) correspondingly. This result is applied to the simple proof of the well-known results on the asymptotic distribution of the zeroes of classical orthogonal polynomials.