The ultraviolet data were taken by IUE on 1994 July 23-24 (i.e. three days prior to the ASCA observation described below.) Due to excessive scattered solar light, we present only the data taken with the short wavelength primary (SWP) camera. These short-wavelength data were taken during one full shift on 1994 July 24 (effective exposure time 8 hours). Data reduction was performed using the ARK software package.
Figure 3: IUE (SWP) data for MCG-6-30-15. Note the weak continuum and the
prominent CIV 
line.
The resulting SWP spectrum is shown in Fig. 3. The source is found to be
weak in the UV compared with other wavebands (apart from radio). The
hypothesis that this UV spectrum is flat and featureless can be rejected
with a high level of confidence: such a hypothesis gives 
for
65 degrees of freedom (dof). The most significant deviation from the
constant model is clearly the CIV resonance emission
line doublet. Modeling this with a Gaussian profile leads to a highly
significant improvements in the goodness of fit ( 
for 3
additional dof). The inferred properties of this line are reported in
Table 1. The CIV line appears to be significantly
broader than the H 
line, and has a FWHM comparable with the
`very-broad' component of the H 
line. The CIV line is also
blueshifted by 
with respect to the broad Balmer lines.
This appears to be a generic feature of high-ionization lines from AGN
(e.g. see Espey et al. 1989). This has been interpreted by some authors as
evidence for a two-component BLR (e.g. Collin-Souffrin et al. 1988).
There may also be an absorption trough to the blueside of the CIV lines. Modeling this with a Gaussian profile leads
to a further improvement in the goodness of fit by 
for 3
additional dof. According to the F-test, this is not a significant
improvement at the 90 per cent level. Thus, we cannot conclusively
determine the reality of this feature and shall not discuss it further.
Formally, the best fit continuum level is 
. It is possible that much of this UV
continuum could be stellar in origin. Thus, this value should be
considered only as an upper limit to the UV continuum flux from the AGN.
High-resolution UV imaging with HST will allow the stellar UV flux to be
almost completely separated from the AGN UV flux, thereby allowing these
two components to be separated. At the time of writing, such UV imaging
has yet to be performed.