2 Sample selection and data

 

 

Table 1: Log of the observations

Name	Other names	mv	z	run1	Exposure	FWHM2
					(minutes)	(arcsec)
Q 0955+326	3C 232	15.8	0.530	1	20	0.36
PG 1001+291		16.0	0.329	1	60	0.24
PG 1012+008		15.6	0.185	1	20	0.26
PKS 1302-102		15.2	0.286	1, 2	20, 48	0.26, 0.42
PG 1402+261		15.5	0.164	1, 2	40, 64	0.26, 0.34
B2 1425+26		15.7	0.366	1, 2	35, 64	0.24, 0.34
Q 1618+177	3C 334	16.0	0.555	1	30	0.36
PG 1700+518		15.1	0.290	1, 2	40, 40	0.26, 0.30
Q 1704+608	3C 351	15.3	0.371	1, 2	40, 64	0.26, 0.43
B2 1721+34		15.4	0.206	1	75	0.24
PG 2112+059		15.6	0.466	1, 2	30, 32	0.24, 0.34
PKS 2128-12		16.2	0.501	2	24a	0.61, 0.48

$\parbox{11cm}{$^1$\space Run 1: may 98 ($H$ -band), Run 2: may 99 ($... ...rs observed close in time to the objects.\\ $^a$\space Also 28 min in $H$ . }$

 

 

Table 2: Characteristics of host-galaxies

Name	z	Companions		Extension		$m_{\rm T}^H$( $m_{\rm T}^K$)	$m_{\rm PSFsub}^H$( $m_{\rm PSFsub}^K$)	mr1/4H( mr1/4K)	Host type
		5 arcsec	10 arcsec	arcsec	kpca
Q 0955+326	0.530			2.2	7.9	14.18	14.25	12.56	E
PG 1001+291	0.329			2.9	8.1	13.87	14.10	14.84	SBa
PG 1012+008	0.185	1	1	3.0	6.3	16.71	16.97	17.19	E-Sa
PKS 1302-102	0.286	2		2.6	7.1	13.45 (12.53)	13.84 (12.89)	13.84 (13.08)	E
PG 1402+261	0.164			3.0	5.3	13.23 (11.91)	14.07 (12.86)	14.07 (11.67)	SBa
B2 1425+26	0.366		2	2.3	7.5	14.26 (13.30)	14.69 (14.01)	15.01 (14.23)	E-Sa
Q 1618+177	0.555	1	2	1.5	5.4	14.68	15.18	15.48
PG 1700+518	0.290	1		2.5	6.7	12.90 (11.75)	13.89 (12.50)	13.89 (12.15)	Sa?
Q 1704+608	0.371		1	2.3	6.9	13.30 (12.21)	14.37 (12.89)	14.30 (12.84)	E
B2 1721+34	0.206		2	1.5	3.2	13.95	14.13	15.07
PG 2112+059	0.466	1	1	2.0	6.6	13.64 (12.64)	14.10 (12.96)	14.71 (13.51)	E?
PKS 2128-12	0.501		1	1.7	6.1	13.42 (12.77)	14.39 (13.33)	14.50 (13.21)
a H0 = 100 km s-1 Mpc-1; $q_{\rm o}$ = 0.5.

In order to use adaptive optics correction quasars were selected such that the nuclei were bright enough to be used as the wavefront reference point source. The sample of radio-quiet quasars were all PG quasars with $m_{\rm b}< 16.5$ and with redshift less than 0.6. The radio-loud objects were selected from 3C, 4C, B2 and PKS catalogues with the same magnitude and z criteria. The final objects observed (see Table 1) were selected based upon the suitability for the observing conditions on the observing runs.

We used the CFHT adaptive optics bonnette (PUEO) and the IR camera KIR on May 1998 (run 1) and May 1999 (run 2). The weather conditions were poor during both runs and the FWHM of the seeing PSF was never better than 0.8 arcsec. The adaptative-optics correction was performed on the QSOs themselves. The quasar was centered successively in the center of the four quadrants of the detector. The exposure time for individual images was two minutes. The background was determined by median-averaging the frames and the flat-field was taken to be the normalized dark-substracted background. The images were then aligned and added. The final images have a typical resolution of FWHM $\sim$ 0.3 arcsec. After each science observation an image of a star with similar magnitude as the QSO was taken in order to determine the PSF and use it to deconvolve the images. Due to rapid variations in the wheather conditions however, it was not always possible to follow this predefined procedure.

A synthetic PSF function, derived from the stellar images was used to deconvolve each of the images. As it was not always possible to apply a standard procedure due to fluctuating seeing conditions, a careful, although somewhat arbitrary choice of the PSF had to be done. In Fig. 2 we show the images of PKS 1700+514 obtained using, for the deconvolution, three different PSFs from stars observed during the same night. These have respectively, FWHM = 0.30, 0.42 and 0.48 arcsec. The initial image of the object has a resolution of FWHM = 0.26 arcsec and the star observed just after the science exposure has FWHM = 0.48 arcsec. It is apparent that the best result is obtained using the star with the FWHM closest to that of the science exposure. Here, we were guided in the exercice by the existence of the HST image by Hines et al. (1999). In general, this illustrates the crucial role played by a careful PSF determination in AO observations.

Results are summarized in Table 2. Columns $\char93 3$ and $\char93 4$ give, respectively, the number of objects (probably companions) found within 5  and 10 arcsec from the quasar down to m_H = 20.5; columns $\char93 5$ and $\char93 6$ give the maximum radial distance (in arcsec and kilo-parsec) to which the host is detected at a significance level of 3$\sigma$ above the background; column $\char93 7$ gives the total magnitude of the object in the H-band and columns $\char93 8$ and $\char93 9$ those of the host-galaxy as derived from the PSF subtraction and profile fitting respectively (see Sect. 4); the assigned morphology, which comes from the 2D brightness distribution and the comparison of the two profile fittings, is given in column $\char93 10$.