We present far-ultraviolet spectroscopy of the cores of the massive cooling flow clusters Abell 1795 and 2597 obtained with FUSE .
As the intracluster gas cools through 3 \times 10 ^ { 5 } K , it should emit strongly in the O vi \ > \lambda \lambda 1032,1038 resonance lines .
We report the detection of O vi \lambda 1032 emission in A2597 , with a line flux of 1.35 \pm 0.35 \times 10 ^ { -15 } erg cm ^ { -2 } s ^ { -1 } , as well as detection of emission from C iii \lambda 977 .
A marginal detection of C iii \lambda 977 emission is also reported for A1795 .
These observations provide evidence for a direct link between the hot ( 10 ^ { 7 } K ) cooling flow gas and the cool ( 10 ^ { 4 } K ) gas in the optical emission line filaments .
Assuming simple cooling flow models , the O vi line flux in A2597 corresponds to a mass deposition rate of \sim 40 M _ { \odot } yr ^ { -1 } within the central 36 kpc .
Emission from O vi \lambda 1032 was not detected in A1795 , with an upper limit of 1.5 \times 10 ^ { -15 } erg cm ^ { -2 } s ^ { -1 } , corresponding to a limit on the mass cooling flow rate of \dot { M } ( 28 { ~ { } kpc } ) < 28 M _ { \odot } yr ^ { -1 } .
We have considered several explanations for the lack of detection of O vi emission in A1795 and the weaker than expected flux in A2597 , including extinction by dust in the outer cluster , and quenching of thermal conduction by magnetic fields .
We conclude that a turbulent mixing model , with some dust extinction , could explain our O vi results while also accounting for the puzzling lack of emission by Fe xvii in cluster cooling flows .