We present extensive optical , infrared , and X-ray observations of the S-type symbiotic low-mass X-ray binary GX 1+4/V2116 Oph , which consists of a 2-min X-ray pulsar accreting from an M6 III giant .
This is the only symbiotic system definitely known to contain a neutron star .
The mean observed spectral type of the X-ray-heated mass donor is M5 III .
The steady interstellar extinction toward the binary ( A _ { V } = 5.0 \pm 0.6 ) contrasts the variable hydrogen column density inferred from X-ray measurements , most likely evidence for a variable stellar wind .
The mass donor is probably near the tip of the first-ascent red giant branch , in which case the system is 3–6 kpc distant and has an X-ray luminosity of \sim 10 ^ { 37 } erg s ^ { -1 } .
It is also possible , though less likely , that the donor star is just beginning its ascent of the asymptotic giant branch , in which case the system is 12–15 kpc distant and has an X-ray luminosity of \sim 10 ^ { 38 } erg s ^ { -1 } .
However , our measured A _ { V } argues against such a large distance .

We show that the dense ( N _ { e } \sim 10 ^ { 9 } cm ^ { -3 } ) emission-line nebula enshrouding the binary is powered by ultraviolet radiation from an accretion disk .
The emission-line spectrum constrains the temperature profile and inner radius of the disk ( and thus the pulsar ’ s magnetic field strength ) , and we mention the implications this has for explaining the accretion torque reversals observed in the pulsar .
We also show that the binary period must be \gtrsim 100 d and is most likely \gtrsim 260 d , making GX 1+4 the only known low-mass X-ray binary with P _ { orb } > 10 d. If the mass donor fills its Roche lobe , the mass transfer rate must be highly super-Eddington , requiring considerable mass loss from the binary .
We discuss the alternative possibility that the accretion disk forms from the slow , dense stellar wind expected from the red giant .