A fluorescent lamp, regardless of shape, is an arc lamp. Works on the same principle as a neon lamp. The electrodes in a neon lamp are close together, the arc can start at a fairly low voltage. The electrodes of a fluorescent are quite a distance apart. They need a kicker to get them going. Such as the very old fixtures that used an external starter. Essentially a thermal timer. Modern technology has eliminated the need for such.
For all practical purposes, an arc is a short circuit. The ionized air around a DC contact is the conductor. Heavy DC contacts have "arc chutes" and "blow out coils" to extinguish arcs. AC usually will normally breaK its' own arc because the arc has to restart every time the line swaps polarity. Only in a very heavy current and short gap will AC maintain an arc. Compare AC to DC welders, for an example.
The gas in a neon or fluorescent lamp, once ionized, is the same. Neon gas for the "neon" lamp, mercury vapor in older fluorescent tubes. And most of the newer ones.
Once the arc is started, as a short circuit, it has no appreciable resistance to limit the current. Hence the series resistor in a neon lamp and the ballast in a fluorescent fixture. It's a current limiter. Same as the ballast in a mercury vapor or sodium HID fixture.
Neon gas fluoresces in the visible spectrum. A fluorescent in the ultra-violet region. A "black light" tube is merely a fluorescent lamp without the phosphors. (white powder) When the powder is struck by UV photons, it glows white in the visible spectrum. Different compositions here will create light of different colours. Hence the DL(daylight), CW(cool white), &c as part of the lamp number.
There are sea stories about playing with fluorescent lamps in the vicinity of high power transmitters. I never bothered to find out if the gas reacts to the RF or the RF caused a current internal to the lamp. Would be an interesting question for someone knowledgable in the physics of the matter.
If you will observe a neon lamp connected across DC, the glow is around the positive plate. On AC, it is around both, the AC swaps polarity (50)60 times a second. Each plate is positive for half of the period. A fluorescent is the same, although you don't see it. The phosphers will tend to light brighter on one end on DC. It was considered routine to swap the lamp end for end on a regular basis in a DC fixture. I don't know if modern ballasts eliminate the need for this or just no one bothers with maintenance anymore.
CFL's (compact fluorescent lamps) are still fluorescent lamps. The use of electronic ballasts has permitted generating curved arcs. But, the physics of the arc are no different from the 8 foot T-12 tubes in a commercial fixture. Merely lower power.
As an aside, it was not unusual to hear the old folks refer to a fluorescent light as a "neon" light fixture, supposedly after neon signs. Transformers / ballasts from neon signs are quite useful for serious high voltage tinkering.