1 Introduction

This work is based on a systematic analysis of images of comet P/Halley collected during its last apparition as part of the International Halley Watch program by the Large Scale Phenomena Network of observatories. A subset of images (531 out of 1439) from The International Halley Watch Atlas of Large-Scale Phenomena (Brandt et al. 1992a), covering the period from 1985 September 17 to 1986 July 06, were selected on the basis of their many conspicuous morphological structures along the cometary tail. They were visually and systematically analysed for the purpose of identification and classification (Voelzke & Matsuura 1998).

Occasionally, the entire or partial plasma tail separates from the comet and drifts away (antisunward), followed by renewal of the plasma tail. The cyclic nature of this phenomenon, called a disconnection event (DE), was noted long ago (Barnard 1899, 1920). Disconnection events (Niedner & Brandt 1979; Jockers 1985; Celnik et al. 1988; Delva et al. 1991; Brandt et al. 1992b; Voelzke & Matsuura 1998) are common. They occur over wide ranges in heliocentric distance, heliospheric latitude, and phases of the solar cycle. Although the mechanism remained unresolved, the solar wind was suspected to play a major role in the phenomena of comet plasma tail DEs. Many ideas were suggested to explain the cyclic phenomena of DEs (Brandt 1990). The current competitive theories to explain the onset of DEs can be grouped into three classes based on the triggering mechanisms: (a) Ion production effects (Wurm & Mammano 1972), (b) Pressure effects (Ip & Mendis 1978; Ip 1980; Jockers 1985; Wegmann 1995; Wegmann 1998), and (c) Magnetic reconnection (Niedner & Brandt 1978; Russell et al. 1986; Brandt et al. 1992c; Yi et al. 1996; Yi et al. 1998; Brandt et al. 1999). In this work the comet P/Halley DEs are interpreted in terms of the location of the magnetic sector boundaries derived from solar magnetic field data on the "source-surface" (Hoeksema et al. 1983) and the expansion of the solar wind.