We present here the analysis of ultraviolet spectra from IUE and an X-ray spectrum from ROSAT PSPC observations of the X-ray weak , far-infrared loud AGN , PG 1126 - 041 ( Mrk 1298 ) .
The first UV spectra taken in June 1992 , simultaneously with ROSAT , show strong absorption lines of NV , CIV and SiIV , extending over a velocity range from -1000 to -5000 km s ^ { -1 } with respect to the corresponding line centre .
Our analysis shows that the Broad Emission Line Region ( BELR ) is , at least partially , covered by the material causing these absorption lines .
In the IUE spectrum taken in Jan. 1995 , the continuum was a factor of two brighter and the UV absorption lines are found to be considerably weaker than in 1992 , but only little variation in the emission line fluxes is found .
With UV spectral indices of \alpha _ { uv } \simeq 1.82 and 1.46 for the 1992 and 1995 data , the far UV spectrum is steep .
Based on the emission line ratios and the broad band spectral energy distribution , we argue that the steepness of the UV spectrum is unlikely to be due to reddening .

The soft X-ray emission in the ROSAT band is weak .
A simple power-law model yields a very poor fit with a UV-to-X-ray spectral index \alpha _ { uvx } = 2.32 .
Highly ionized ( warm ) absorption is suggested by the ROSAT data .
After correcting for a warm absorber , the optical to X-ray spectral slope is close to the average of \alpha _ { uvx } \simeq 1.67 for radio quiet quasars .

From photoionization calculations we find : ( 1 ) A single zone absorption model can not explain simultaneously the UV absorption lines and the ionized X-ray absorption if metal abundances are solar .
Furthermore , in order to be consistent with the equivalent width of the observed Ly \alpha absorption line , the turbulent velocity of the warm absorber must be less than 190 km s ^ { -1 } , which imposes serious constraints on a disk wind model .
( 2 ) The UV absorption lines and their variability can not be explained by a single zone model with solar abundances and the large variability in the absorption lines suggest that CIV and NV absorption lines are not severely saturated .
( 3 ) The absorption of the ionizing continuum by warm material strongly affects the emission line spectrum .