We used FUSE to observe ultraviolet emission from diffuse O VI in the hot gas in the Galactic halo . By comparing our result with another , nearby observation blocked by an opaque cloud at a distance of 230 pc , we could subtract off the contribution from the Local Bubble , leading to an apparent halo intensity of I _ { OVI } = 4680 ^ { +570 } _ { -660 } photons cm ^ { -2 } s ^ { -1 } sr ^ { -1 } . A correction for foreground extinction leads to an intrinsic intensity that could be as much as twice this value . Assuming T \sim 3 \times 10 ^ { 5 } K , we conclude that the electron density , n _ { e } , is 0.01 - 0.02 cm ^ { -3 } , the thermal pressure , p / k , is 7000 - 10 , 000 { cm } ^ { -3 } K , and that the hot gas is spread over a length of 50-70 pc , implying a small filling factor for O VI -rich gas . ROSAT observations of emission at 1/4 keV in the same direction indicate that the X-rays are weaker by a factor of 1.1 to 4.7 , depending on the foreground extinction . Simulated supernova remnants evolving in low density gas have similar O VI to X-ray ratios when the remnant plasma is approaching collisional ioinizational equilibrium and the physical structures are approaching dynamical “ middle age ” . Alternatively , the plasma can be described by a temperature power-law . Assuming that the material is approximately isobaric and the length scales according to T ^ { \beta } d \ln T , we find \beta = 1.5 \pm 0.6 and an upper temperature cutoff of 10 ^ { 6.6 ( +0.3 , -0.2 ) } K. The radiative cooling rate for the hot gas , including that which is too hot to hold O VI , is 6 \times 10 ^ { 38 } { erg~ { } s } ^ { -1 } { kpc } ^ { -2 } . This rate implies that \sim 70 \% of the energy produced in the disk and halo by SN and pre-SN winds is radiated by the hot gas in the halo .