Scatter from field-aligned irregularities associated with equatorial spread F has been studied using the powerful 50 MHz radar at Jicamarca. We find that these irregularities, which have a wavelength of 3 meters, first, at times can be generated anywhere in the F region, no matter with what vertical velocity the region is moving, second, have growth times of a few seconds or less, third, move at velocities comparable to that of the F layer, and fourth, may attain strengths of perhaps 107–108 times the background thermal level. There are also indications that there may be two types of irregularities, classified roughly as strong and weak, and that there may be a minimum threshold altitude that the bottom of the F layer must be above before the irregularities are first generated. These and other observed characteristics of the irregularities seem to rule out currently popular theories of their origin (the vertical drift instability suggested first by Martyn, mechanisms based on coupling between the E and F regions, gravitational instabilities, generation by hydromagnetic waves, and others) and several other well known types of plasma instability. Gradients of density and drift velocity remain as possible sources of instability, but at the moment we can only speculate as to what may ultimately prove to be the correct explanation.