It is possible to ``observe'' the model from any arbitrary angle
to deduce quantities to be compared with observations.
Figure 2 shows contours of ``surface brightness'' towards the inner galaxy
from the point marked in the last frame. The particles were projected into a
raster of 
bins and weighted by the inverse square of their
distance from the observer, except those closer than half the distance to the
centre were discarded. The viewing angle is 30
to the bar major
axis; if this angle is increased, the bulge acquires too much of a peanut
shape to be consistent with the extinction corrected COBE image (Weiland
et al., this meeting), while if it is reduced, the asymmetry between the two
sides becomes more pronounced. The distance from the centre was chosen in
order that the box-like feature in the third contour down from the peak
subtends an angle of approximately 10. (Equating this distance to
8 kpc and setting the velocity unit at 300 km/s, then the time unit of
the simulation is 4.33 Myr and the total mass is 
.) The resemblance to the Weiland et al. COBE photometry is quite
striking, although the asymmetry in the 
extent of the bar between the
near and far sides is perhaps too strong.
Figure 3: The line-of-sight components (full drawn) and perpendicular
components (dashed) over a range of longitudes at fixed latitude. All
particles within 2 kpc of the galactic centre distance contribute
equally and data from above and below the plane combined to improve
statistics. The bottom pair of curves are for 
, higher pairs are
for values increasing by 2 and offset vertically by
100 km/s each time for clarity.