Grayline is caused by the D-layer, which dissapears during sunset. On the other side of the line, at sunrise, it is only beginning to build. This can create incredibly efficient propagation along the grayline, north and south. Lengths of openings depend mainly on frequency; the lower bands (40, 80, 160) can experience grayline openings of only a few moments in length, whereas the high bands (10, 15, 20) can produce openings of thirty minutes or more. Using the darkness path, western stations are louder in the morning, during the sunrise period, and eastern stations are louder in the evening, during sunset.
For stations here in New England, the grayline can produce terrific openings into Asia, especially on 15 and 20. 10 meters will also experience very nice openings, as long as normal band conditions are propagatory. For amateurs in this area (generally in the Northeast), they should point their beams NW (about 330 degrees), since JA is not towards EU at all! In fact, if you cannot point your beam NW, you are better off pointing it W than towards EU!
Signals propagating along grayline paths tend to be fluttery or even "drown out", which usually leads to their confusion with longpath signals. Since these paths tend to be vertical (north and south) in nature, they can also be confused for transequatorial modes, in the case of openings to the south. Though most paths are vertically-oriented along the grayline corridor, they are usually far too long to be transequatorial, but far too short to be longpath.
Still need those little-heard Asian prefixes? Now, there's a solution! Spin that beam, and watch the sunset from your shack. When I was younger, I would sit on 15m at dusk every night and work an absolute string of JA's and BY's. Even in the down-years of the cycle, when the A and K indices are too low for much else, grayline is still viable.
Web links, maps, and charts:
Grayline map with maximum useable frequencies
Map using The Living Earth imaging, from John Walker
Shows terminator along entire globe