scientific article; zbMATH DE number 3897960

From MaRDI portal
Publication:3676921





It is wellknown that a Markov process satisfying the relation \[ (1)\quad \lim (1/t)P_ t(x,U^ c)=0,\quad U\quad open,\quad x\in U, \] is continuous. \textit{D. Ray} [Trans. Am. Math. Soc. 82, 452-493 (1956; Zbl 0072.351)] proved that, within the class of Feller processes on the real line, the above relation characterises the continuous ones. He also raised the question whether the same characterisation would hold on more general locally compact spaces. The present note gives a counterexample in the plane, which shows that this is not true. However the author succeeded in corollary 4.12 from page 95 of his book Local operators and Markov processes (1980; Zbl 0446.60067) to characterise the continuous processes in terms of resolvents. The note also shows that there exists a strict potential which is not hyperstrict, giving so a negative answer to a question posed by \textit{C. Constantinescu} and \textit{A. Cornea} in axiomatic potential theory [Potential theory on harmonic spaces. (1972; Zbl 0248.31011)].











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3676921)