\documentstyle[12pt,paasms4,psfig,astrobib]{preprint}		
\hoffset=-.1in
\input pub.sty
\include{symbols}

\begin{document}

\title{ECLIPSE MAPPING OF THE ACCRETION DISK WIND IN\\
~\\
THE CATACLYSMIC VARIABLE UX~UMa\thanks{Based in part on observations with the
NASA/ESA {\em Hubble Space Telescope}, obtained at the Space Telescope Science
Institute, which is operated by the Association of Universities for
Research in Astronomy, Inc., under NASA contract NAS5-26555.}}

\author{Christian Knigge\\
\\
Space Telescope Science Institute\\
\\
3700 San Martin Drive\\
\\
Baltimore, MD 21218\\
\\
knigge@stsci.edu\\
\and
Janet E.\ Drew\\
\\
Imperial College, Blackett Laboratory\\
\\
Prince Consort Road\\
\\
London SW7 2BZ, UK\\
\\
j.drew@ic.ac.uk}

\tobe{1 September 1997}{Ap.~J.}
\recacc{29 January 1997}{25 March 1997}
\maketitle

\section*{ABSTRACT}

We present the results of an effort to model recent HST eclipse
observations of the C~{\sc iv}~1550~\AA\ resonance line in the
high-inclination nova-like cataclysmic variable UX~UMa. Assuming 
this line to be predominantly wind-formed, we are able to roughly
reproduce the shapes and strengths of the observed line profiles 
(both away from and during eclipse) within the framework of a simple 
kinematic model in which the outflow is described as a rotating,
biconical accretion disk wind. Our preferred model also
matches most of the detailed behaviour of different parts of the C~{\sc
iv} line profile (blue wing, line centre and red wing) as a function of
orbital phase. 

The most important result of our modeling is that it strongly suggests the 
presence of a high column ($N_H \sim10^{23}$~cm$^{-2}$), relatively 
dense ($n_e \simeq 4 \times 10^{12}$~cm$^{-3}$) and only slowly-outflowing 
($v_{poloidal} \ll v_{escape}$) transition region between the disk
photosphere and the fast-moving wind. We also find that the outflow from
UX~UMa is probably only moderately collimated (the outer opening angle of our
preferred model is $\theta_{max} \simeq 65^{\rm o}$). Finally, the rotational
disturbance seen in the data is fit reasonably well by our model, 
in which the rotational component of motion in the wind is fixed by 
conserving the specific angular momentum of the outflowing
material. The implications of these results for dynamical models of
disk-driven winds are discussed.