1 Introduction

It is well known that the phase and amplitude errors in the complex visibilities observed with a radio interferometer array can be effectively removed making use of closure techniques [4, (Jennison 1958;] [13, Rogers et al. 1974;] [12, Readhead & Wilkinson 1978)] and redundancy in the length and orientation of baseline vectors [3, (Ishiguro 1974;] [6, Noordam & de Bruyn 1982;] [14, Wieringa 1992;] [5, Nakajima et al. 1994).] These techniques make use of the fact that most of the errors can be ascribed to individual antennas (see [7, Pearson & Readhead 1984] for the details). In this paper a scheme based on these techniques is described for eliminating the phase errors in a T-shaped interferometer array particularly those in which the tilting of the antenna beam is done through introduction of phase/delay gradients across the array.