I got 28AWG titanium wire, but unfortunately it did not seem to perform very well.
Out of the box it appeared like a rather dull metal with no obvious indication that it was composed of titanium. The wire is easy to bend.
However upon testing, it clearly behaves differently than the wires tested so far. Under cathodic electrolytic action, it tends to turn into a brittle blue-black material. This is probably titanium hydride, but some non-stoichiometric oxide may be present too.
The ball shape is formed because when it gets hot the wire appears to start combusting at very high temperatures. This process cannot be stopped easily by immersion in water (i.e. it will combust under water).
The remaining surface is still conductive, at least the voltages used. Sometimes it turns into a light green/gray, irregularly-shaped compound that looks like this. It's conductive in this form:
Other times this combustion eats up large portions of the wire at a time, like here, but strangely slowly compared to copper or molybdenum. Note the 'ghostly' appearance of the bottom portion of the wire (EDIT: perhaps it was a low-current arc?):
The "combusted" wire can be used on the top of the water level for a kind of plasma heating reaction (from atmospheric glow discharge), but it cannot be immersed too much in the electrolyte without immediate quenching. It appears to slowly vaporize and form residues (possibly TiO2) in its immediate surrounding when kept at high temperature like this.
Overall, in any case, performance for the plasma electrolysis experiments described in this thread seems rather poor, perhaps even less performing than the Kanthal wire I got a while ago. Keeping the plasma going on for more than a few millimeters under the water level does not seem feasible. Since this behavior is similar to that of the Kanthal wire, it's probably mainly the fault of the oxide layer formed on the surface, and possibly the high resistivity.
No change observed with the Geiger counter. RF emissions were rather low too, together with the current instability effect observed earlier, also after adding about 1g acetone and 1 ml 0.1M KOH solution as for the previous tests.