International Journal of Mathematics and Mathematical Sciences
Volume 16 (1993), Issue 4, Pages 749-754
doi:10.1155/S0161171293000936

On defining the generalized functions δα(z) and δn(x)

E.K. Koh and C.K. Li

Department of Mathematics and Statistics, University of Regina, Regina S4S 0A2, Canada

Abstract

In a previous paper (see [5]), we applied a fixed δ-sequence and neutrix limit due to Van der Corput to give meaning to distributions δk and (δ)k for k(0,1) and k=2,3,. In this paper, we choose a fixed analytic branch such that zα(π<argzπ) is an analytic single-valued function and define δα(z) on a suitable function space Ia. We show that δα(z)Ia. Similar results on (δ(m)(z))α are obtained. Finally, we use the Hilbert integral φ(z)=1πi+φ(t)tzdt where φ(t)D(R), to redefine δn(x) as a boundary value of δn(zi ϵ ). The definition of δn(x) is independent of the choice of δ-sequence.