Authors have observed an unusual phenomenon that the collector current in an Extractor vacuum gauge, which several times experienced a bakeout of 570 K, about 70 h, becomes negative at pressures in or near the XHV region. To investigate the causes, the temperature dependence of the collector current, the electromotive force (e.m.f.), and the insulation resistance of the collector electrode for both an "unusual" and a new gauge were measured. As a result, their dependence for the unusual gauge showed strikingly different behaviors from those for the new gauge. These results suggest that some new structure is formed in the gauge head by several bakeouts. Also it was found that the new gauge shows a positive larger e.m.f., but the current due to this e.m.f. was very small because the collector electrode has a very high insulation resistance.