Found this,does this sound correct, the part about the float contacts crossing over in an "X"?

The retarded part is that there are two rods and not one and the design of the float, but I digress....

The chewing tobacco can-shaped float has an oval space in the middle into which the two rods fit in and the float rides up and down upon. There is a little clip which holds the chinsy little contact set up. The contact piece is basicaly a rectangular back piece with 2 thin arms extending from either side with almost spalula end-shaped rectangular contacts on the end of them. The little thin arms actually cross over one another in an "X" and the contact pads ride on the outside edge of each rod. If you were looking at the float with contacts in and rods in place, with the contacts facing upwards (it's mounted contact clip side-down when actually in the cylinder), the contact pads would ride at approximately the 4o'clock and 7o'clock on the rods. Theoretically, it's a good idea, but put it into practice for a hundred or so thousand and you begin to get real world problems. First off, the system relies on the fact that the spring tension in the little contact clippy thing remains constant so as to ride the windings properly. Second, the contact pads wear and actually develop grooves where they ride on the rods, affecting the conductivity. Third, we're dealing with a contact piece that, when jostled just right, can have a contact arm/pad slip from the outside of the rod to the inside, thus creating a zero reading on the guage--and this usually happens when the pump is changed, just due to the twising and jiggling needed to get the pump out---which is what happened to my unit from the awesome "mechanics" who did the work when the previous owner had the pump replaced. Fourth, it's just not designed to be serviced.