We have determined the total flux densities of the entire sources and -
where possible - of individual components. All maps, except the 2.7-GHz
data, have been smoothed to a common beam size of 
(determined by the resolution at 4.8 GHz). In the case of
3C 326 the final beam size in declination is defined by the original
resolution of the 326-MHz map (
). At 2.7 GHz
we took the original beam of 
. Because of this
lower resolution it was difficult in some cases to integrate across the same
areas. The larger beam may also lead to higher flux densities because of
confusion. Their effect will disappear in most cases when determining
spectral indices pixel by pixel which will be reported in a forthcoming
paper. The 609-MHz data are affected by missing-spacing effects, which
implies that the flux densities are lower limits.
As already mentioned, the 2.7- and 4.8-GHz maps had to be cut in some cases. When that happened we did not give flux densities. The error calculation includes errors introduced by zero-level uncertainty, errors coming from the map noise, and calibration errors. Since systematic errors such as missing flux due to missing spacings, variable baselines, or different noise levels across the map cannot be included, the error ranges might be underestimated in some cases. Therefore, we also show the flux densities in graphic form and comment on the spectra where necessary. For the sake of clarity we do not show any spectra of the background sources (bgs). All individual areas used to integrate the various component flux densities, their numbers given in Tables 7 (click here)-11 (click here), are shown shaded in the finding charts (Fig. 5 (click here)).
The spectrum (
) of the entire source is
obviously dominated by the
core which reveals a significant flattening towards higher frequencies
(
; 
).
Here 
is the spectral index between 326 and 609 MHz,
is determined between 4.8 and 10.6 GHz.
We have determined the spectrum of the south-eastern lobe in both the southern
brighter and the northern fainter part. The spectrum of the brighter part is straight,
with 
. The flux densities of the weak part are
much more scattered, but the steeper spectral index of
is obvious.
Table 11: Integrated flux densities of NGC 6251
If the 2.7-GHz point is excluded, the bright western lobe has an almost
straight spectrum, with 
. The determination of a
spectrum of the weak part of the western lobe is not possible since the flux
densities are too strongly distorted.
The spectrum of the western hot spot flattens towards high frequencies
(
; 
).
The counter-jet region has a typical jet spectrum (
) up
to 4.8 GHz. The bright knot within the jet has a spectral index of
.
The possible background source in the counter-jet area possesses a spectrum of
, with a slight trend for a steepening
towards higher frequencies.
The spectrum of the huge backlobe of NGC 315 is still not well determined.
The appearance of a typical spectrum with 
between 326
MHz and 4.8 GHz is in contrast to the excess at 10.6 GHz, which has already been
noticed by Klein et al. (1994), and is still significant even
after CLEANing.
The total spectrum is obviously dominated by the influence of the very bright
hot spot within the eastern lobe, which shows a flattening of the spectral index
towards higher frequencies (
;
). The spectra of the western lobe and the
core are almost
straight, with 
and 
,
respectively. The slight relative
excess of the 2.7-GHz flux density can be explained by confusion of the core
region when observed with the 
beam.
The spectrum of the entire source shows a deficit of the 609-MHz value,
which can be explained by missing spacings of the interferometer.
The core spectrum shows excessive 2.7- and 4.8-GHz values, most likely because
of confusion of the western lobe, and possibly owing to
CLEAN artifacts. The overall core spectrum between 326 MHz and 10.6 GHz is
.
The eastern lobe spectrum reveals similar problems. The 609-MHz flux density is
too low relative to the other values, which can again be explained by missing
zero spacings. The 2.7-GHz and maybe also the 4.8-GHz point suffer from
confusion with the core emission. If the 4.8-GHz point is regarded as real,
a slight convex curvature of 
is indicated. Between 326 MHz
and 10.6 GHz a spectral index of 
has been derived.
The western lobe spectrum (
) has also been determined
using the measurements at the lowest and highest frequency only, since the 2.7-GHz flux
density is too high (confusion) and the 4.8-GHz value is uncertain because of an
artificial depression in this part of the map.
The western hot spot again reveals a slight flattening (
; 
). The measurements in the
eastern hot spot area and of the background source west of the hot spot are
too uncertain to determine the overall spectrum, owing to confusion.
All spectra determined here show a clear steepening towards higher frequencies. The low- and high-frequency spectral indices have been compiled in Table 12 (click here).
Table 12: Spectral indices of 3C 326 (Errors are in all cases 
)
At 2.7 and 4.8 GHz, there are no reliable flux densities available for the two
lobes, nor for the entire source, as the maps had to be restricted in size at these frequencies.
The spectral indices between 326 MHz and 10.6 GHz are 
in the western lobe, and 
for the entire source.
In the case of the eastern lobe we can only determine a low-frequency spectral
index of 
. The core spectrum is almost straight, with a
slight indication
of a flattening towards high frequencies. The (straight) spectral index is
.
The hot spot spectra are straight within the errors, with a large
difference in the spectral indices between the eastern (
) and the
western one (
). In the latter the influence of confusion
is high because of the close neighbourhood of bright sources.