In the light of the postings above, I'm just going to confirm that wireless breadboard is absolutely insidious when it comes to constructing resonant circuits: I've largely ignored the advice above, and I've continued to play around with tank circuits on breadboard, which I've been driving from a sig. gen. through a fixed resistor (I've been swapping in different values to see how it affects Q). These have been behaving largely as expected.
This afternoon, I swapped the fixed resistor for a variable, and I was baffled to see that I was now getting voltage gain from the top of the tank. It took me over an hour to figure out that in fact I'd managed to construct a series LC circuit, where the series C was due to the capacitance between the legs of the variable resistor which were in adjacent rows of the board - in fact, the circuit worked unchanged if I removed the resistor completely, and merely relied on the capacitive coupling between the rows.
This is, on the one hand, painful, and on the other, very instructive.