3 Results

3.1 Identifications

Similar to the procedure outlined in Paper I, we used the likelihood ratio

$$R \quad = \quad \sqrt{ \frac{\Delta\alpha^2}{\sigma^2_\alpha} + \frac{\Delta\delta^2}{\sigma^2_\delta} }$$ (1)

as described by De Ruiter et al. (1977) to quantify the correctness of a particular identification. $\Delta\alpha$ and $\Delta\delta$ are measured offsets in RA and Dec between the optical and radio positions, $\sigma_\alpha^2$ and $\sigma_\delta^2$ are the sums of the squared 1$\sigma$ errors in the optical and radio positions. Note that in most cases the radio accuracy is much higher than in the optical (cf. Col. 4 in Table 1), so $\sigma_\alpha^2$ and $\sigma_\delta^2$ are dominated by the optical error in these cases. We estimate our optical accuracy to be 0.5 $^{\prime\prime}$ as in all cases we were able to identify HST guide stars within the field of view. An R-value less than three indicates a chance less than 1% to miss a true identification (assuming that the possible counterpart is the object closest to the radio position). The probability that a true optical counterpart has an R-value larger than some R₀ is given by $P(R > R_0) = {\rm e}^{-R^2_0 / 2}$. Most of our identifications have R-values smaller than 2, and can be assumed correct. Besides, some of the higher R-values, for instance in the case of 1601-222 or 1648+015, do not necessarily indicate a misidentification. Their high R-value may be due to overestimates of the accuracy, which will artificially inflate the R-value. The identification results are listed in Table 2, and finding charts for the sources are presented in Fig. 4. The center of the cross indicates the radio position, and the angular size of the cross hairs is fixed at 20 $^{\prime\prime}$, independent of the plate scale of any particular observation.

 

 

Table 2: Positions and magnitudes of identifications

	Optical position (2000.0)
Name	h	m	s		$^\circ$	$^\prime$	$^{\prime\prime}$	R-value	ID	maga
0437-454	04	39	00.84		-45	22	22.3	0.3	G	19.0 I
0742+103	07	45	33.17		10	11	15.3	4.4	G	$\sim24$
0802+212	08	05	38.61		21	06	51.6	0.9	G	22.5 R
0904+039	09	06	40.98		03	42	42.3	0.8	G	22.1 I
1433-040	14	35	40.08		-04	14	54.5	1.3	G	18.3 R
1540-077b	15	43	01.64		-07	57	06.8	0.4	G	17.8 R
1601-222	16	04	01.41		-22	23	42.6	2.9	G	19.3 R
1648+015	16	51	03.72		01	29	21.4	3.9	G	21.4 R
1732+094	17	34	58.42		09	26	59.5	1.1	G	21.3 R
1815-553	18	19	45.40		-55	21	20.4	0.6	G	22.0 V
1942+722	19	41	27.29		72	21	43.7	3.6	G	17.64 $\pm$ 0.05 K
2121-014	21	23	39.26		-01	12	34.2	0.9	G	18.18 $\pm$ 0.08 K
2128+048	21	30	32.94		05	02	17.3	1.8	G	17.47 $\pm$ 0.04 K
2322-040	23	25	10.25		-03	44	47.3	0.6	G	17.46 $\pm$ 0.04 K

^a Indicated filters are: Johnson V, I, R, and the near-IR K-band.

^b Some confusion reigned about the exact radio position of this source. The optical position listed here is the correct identification (cf. notes on this object).

3.2 Magnitudes

We estimate the photometric accuracy for the optical bands to be on the order of 0.1 magnitude for the brighter sources (m 21), increasing to $\sim$0.4 mag for the faint end of the sample (m$\sim$22.5). None of the nights were photometric, and most of the sources are not properly exposed for accurate photometry. In case of an empty field (e.g. 1045+019), the optical counterpart must be fainter than the limiting magnitude of the CCD frames. A conservative estimate of this limit is $\sim$23 (in any of the bands). The photometry results can be found in the last column of Table 2.

3.3 Spectra and redshifts

For 22 of the 29 sources, low dispersion spectra were taken with the sole objective of assessing redshifts. The typical $\sim$30 Å resolution is sufficient to resolve the [O III] 4959/5007 Å doublet, one of the most prominent emission features in optical spectra of radio galaxies. Furthermore, this resolution is a reasonable trade-off between signal-per-pixel and spectral resolution. The results are listed in Table 3. The spectra with positive line identifications are displayed in Figs. 1 and 2.

The following sources remain without proper redshift determination:

0018+729, 0159+839, 0316+161, 0437-454, 0602+780, 0703+468, 0742+103, 0802+212, 0904+039, 0914+114, 1045+019, 1433-040, 1601-222, 1648+015, 1732+094, 1815-553, 1942+772, 2322-040