VLA observations were carried out in 1994 and 1995 at 1.36, 1.66, 4.8 GHz in A, B, and D configurations. WSRT observations took place in Feb. 1994 at 1.4 GHz. In Table 1 we report the journal of these observations.

**Table 1:** Journal of the observations: Col. [1] date of the observations, Col. [2] array (A, B, D, for the VLA configurations, W for WSRT), Col. [3] observing frequency, Col. [4] bandwidth
date	array	$\nu$ (GHz)	$\Delta\nu$ (MHz)
18/04/94	A	1.36	50
	A	1.66	25
29/04/94	A	1.36	50
	A	1.66	25
04/07/94	B	1.36	50
	B	1.66	25
	B	4.88	50 $\times$ 2 IFs
22/01/95	D	1.36	50
	D	1.66	25
21/02/95	D	1.36	50
	D	1.66	25
02/94	W	1.40	40

The VLA data reduction has been performed using the Astronomical Image Processing System (AIPS) developed at the National Radio Astronomy Observatory (NRAO). After a standard calibration we performed several iterations of imaging and phase self-calibration and one final iteration of phase and amplitude self-calibration.

We observed with the VLA in A and B configuration in order to have good sensitivity to radio emitting regions with angular sizes up to 120 arcsec; the high resolution provided by the A-array data allowed a good measure of the core emission and therefore we managed to have an accurate determination of the extended emission flux density. We also planned to combine the data at 1.36 and 1.66 GHz to improve the uv-coverage and increase the sensitivity, but the latter frequency was generally affected by Radio Frequency Interferences (RFI) and we could not pursue our goal. Short B array observations were carried out at 5 GHz to study and compare the arcsecond scale structure detected in the L band. Finally, a few sources were observed in the D array to search for arcminute scale emission (like in 1807+698).

At 1.36 GHz the images were obtained with a multi-field clean to remove the contribution of strong sources in the field. The rms noise in the final images is typically in the range 0.07-0.18 mJy/beam, close to the expected thermal noise, with the exception of the sources with declination lower than $-20^\circ$ . The dynamic range (peak/noise) is between 1000-9000, with typical values around 5000. The images at 1.66 GHz have been only used to verify the results obtained at 1.36 GHz.

We found it useful to combine L band data taken with different configurations only for a few sources in order to increase the sensitivity to the extended emission while maintaining a good resolution. Given that the flux density of the unresolved core may vary significantly between the epochs of the observations in the A and B configurations, we had to normalize the core flux density before combining the different data sets of the same source.

From the A and B array observations we obtained two subsets, each containing only the visibilities with common uv-coverage. Then we produced an image for each of these subsets and determined the peak flux densities. We subtracted from the original data set with the strongest core a pointlike component with intensity corresponding to the difference between the peak fluxes. Finally we combined the two complete data sets. Only for 1807+698 we also added the data in the D configuration, using the same technique.

2.2 The WSRT data

The WSRT observations were carried out to search for arcminute scale extended emission (like for VLA D-Array data).

The total observing time for the observations at the WSRT was 24 hours. Each source was observed for typically five 15-min snapshots, inclusive of slewing times, in a range of HA in order to obtain an acceptable uv-coverage. The typical resolution was about $13''\times13''$ cos $\delta$ and the rms noise in the image plane was between 0.15 and 0.3 mJy/beam. The data reduction has been done using the NEWSTAR package that allows redundancy self-calibration. The flux density scale has been referred to Baars et al. (1977) by means of the observations of 3C 286, whose flux density was assumed to be 14.77 Jy at 1.40 GHz.

The NEWSTAR package allows for the removal of unresolved components (i.e. components with size much smaller than the observing beam) in the imaging process. This has been used to remove the nucleus of each BL Lac object in order to estimate the total flux density of the extended emission.

2.3 Additional images

We also searched for images of the 1 Jy BL Lac sources in the NRAO VLA Sky Survey (NVSS, Condon et al. 1998) and in the FIRST (Becker et al. 1995) survey. Most sources appeared pointlike, except 1514-241 on the NVSS and 0828+493 and 1418+546 on the FIRST. These three images are presented here along with the images obtained from our data, since they show more details than those available in the literature.

**Table 2:** Image parameters. We do not present images for the sources marked with an asterisk because they are unresolved or because the extended structure in our images does not improve the information in the literature. Column [1] IAU name, Col. [2] array (A, B and D, are for the VLA configurations, W is for the WSRT). "a'' and "b'' means that the image is from the FIRST or NVSS, respectively, Col. [3] frequency of the observations, Col. [4] beam major and minor axes, Col. [5] position angle of the restoring beam, Col. [6] rms noise on the image plane, Col. [7] peak flux density on the image. When the core has been subtracted, the flux density not restored in the image is shown in parentheses
[1]	[2]	[3]	[4]	[5]	[6]	[7]
name	array	$\nu$	beam	PA	rms	peak
		(GHz)	(arcsec)	( $^{\rm o}$ )	(mJy/beam)	(mJy/beam)
0048-097	A+B	1.36	2.4 $\times$ 1.1	46	0.08	511
	B	4.88	2.2 $\times$ 1.9	-60	0.15	982
0118-272	A	1.36	3.0 $\times$ 1.2	-21	0.10	742
	B	4.88	5.3 $\times$ 1.4	36	0.15	663
0138-097	A+B	1.36	2.3 $\times$ 1.5	42	0.07	541
	B	4.88	4.3 $\times$ 1.7	-53	0.15	695
0426-380	A	1.36	4.4 $\times$ 1.0	23	0.10	624
	B	4.88	5.3 $\times$ 1.4	15	0.15	1371
$0454{+}844\ast$	W	1.40	9.2 $\times$ 26.8	57	0.20	1 (310)
0537-441	A	1.36	5.3 $\times$ 1.2	1	0.50	3010
	B	4.88	7.1 $\times$ 1.3	-3	1.80	5138
$0716{+}714\ast$	W	1.40	17.5 $\times$ 11.0	0	0.30	83 (280)
$0814{+}425\ast$	W	1.40	19.5 $\times$ 13.5	22	0.25	10 (1099)
$0820{+}225\ast$	W	1.40	33.7 $\times$ 13.8	12	0.30	81 (1988)
$0823{+}033\ast$	D	1.36	54.0 $\times$ 43.0	-52	0.15	1479
0828+493	B $^{\bf a}$	1.39	5.4 $\times$ 5.4	0	0.13	355
$\ast$	W	1.39	18.3 $\times$ 13.1	52	0.3	10 (288)
$0851{+}202\ast$	W	1.40	39.4 $\times$ 12.5	3	0.30	2 (1143)
0954+658	A	1.36	1.1 $\times$ 1.0	34	0.08	597
	B	4.88	2.8 $\times$ 1.3	49	0.15	523
1144-379	A	1.36	3.6 $\times$ 1.1	-9	0.5	21 (1892)
1147+245	A+B	1.36	1.3 $\times$ 1.3	0	0.07	796
	B	4.88	2.6 $\times$ 1.9	82	0.10	845
1308+326	B	1.46	4.3 $\times$ 4.3	0	0.15	859
1418+546	B $^{\bf a}$	1.40	5.4 $\times$ 5.4	0	0.15	566
	W	1.40	17.0 $\times$ 12.0	14	0.25	19 (710)
	D	1.36	68.0 $\times$ 32.0	-69	0.07	815
1514-241	A+B	1.36	3.0 $\times$ 2.0	28	0.15	1606
	B	4.88	3.0 $\times$ 1.5	-50	0.17	2918
	D $^{\bf b}$	1.40	45 $\times$ 45	0	0.40	1993
$1519{-}273\ast$	A	1.36	3.4 $\times$ 1.4	-30	0.2	1690
$\ast$	B	1.36	9.3 $\times$ 3.2	33	0.10	883
1652+398	B	1.36	3.5 $\times$ 3.0	-84	0.13	1383
	B	4.88	1.6 $\times$ 1.3	-76	0.18	1320
$1749{+}701\ast$	W	1.40	18.3 $\times$ 11.6	-61	0.20	1 (650)
1803+784	B	1.36	9.3 $\times$ 4.9	-11	0.07	1761
	B	4.88	2.5 $\times$ 1.2	-13	0.30	2214
1807+698	A	1.36	3.2 $\times$ 1.5	-49	0.15	1136
	A+B+D	1.36	7.0 $\times$ 5.5	-22	0.10	1284
	B	4.88	2.5 $\times$ 1.3	-9	0.15	1507
$1823{+}568\ast$	W	1.40	30.0 $\times$ 10.0	-53	0.30	50 (1190)
2007+777	W	1.40	24.0 $\times$ 10.0	-47	0.25	17 (1095)
2131-021	A	1.36	1.6 $\times$ 1.5	1	0.20	1291
	B	4.88	1.9 $\times$ 1.4	-22	0.20	1669
2240-260	A	1.36	2.0 $\times$ 1.0	6	0.08	813
	B	4.88	3.7 $\times$ 1.3	-29	0.20	805
$2254{+}074\ast$	D	1.36	47.0 $\times$ 42.0	4	0.15	284

2 Observation and data reduction

2.1 The VLA data

2.2 The WSRT data

2.3 Additional images