During the solar minimum, when the Sun is at its least active, the solar wind(1,2) is observed at high latitudes as a predominantly fast (more than 500 kilometres per second), highly Alfvenic rarefied stream of plasma originating from deep within coronal holes. Closer to the ecliptic plane, the solar wind is interspersed with a more variable slow wind(3) of less than 500 kilometres per second. The precise origins of the slow wind streams are less certain(4); theories and observations suggest that they may originate at the tips of helmet streamers(5,6), from interchange reconnection near coronal hole boundaries(7,8), or within coronal holes with highly diverging magnetic fields(9,10). The heating mechanism required to drive the solar wind is also unresolved, although candidate mechanisms include Alfven-wave turbulence(11,12), heating by reconnection in nanoflares(13), ion cyclotron wave heating(14) and acceleration by thermal gradients1. At a distance of one astronomical unit, the wind is mixed and evolved, and therefore much of the diagnostic structure of these sources and processes has been lost. Here we present observations from the Parker Solar Probe(15) at 36 to 54 solar radii that show evidence of slow Alfvenic solar wind emerging from a small equatorial coronal hole. The measured magnetic field exhibits patches of large, intermittent reversals that are associated with jets of plasma and enhanced Poynting flux and that are interspersed in a smoother and less turbulent flow with a near-radial magnetic field. Furthermore, plasma-wave measurements suggest the existence of electron and ion velocity-space micro-instabilities(10,16) that are associated with plasma heating and thermalization processes. Our measurements suggest that there is an impulsive mechanism associated with solar-wind energization and that micro-instabilities play a part in heating, and we provide evidence that low-latitude coronal holes are a key source of the slow solar wind.