3 Results

The results of our observations are summarized in Tables 2-4. Tables 2 and 3 contain the information for radio-loud AGNs, while the results for radio-quiet objects are displayed in Table 4. In these tables Col. 1 lists the object name; Col. 2, the AGN classification; Col. 3, the date of observation; Col. 4, the observational error $\sigma$ obtained from the standard deviation of the comparison differential lightcurve; Col. 5, the duration of each intranight observation. Column 6 gives the classification of each lightcurve according to a variable (V) - nonvariable (NV) scheme. We have adopted the same 99%-confidence criterion used by Jang & Miller (1997) to distinguish between V and NV sources. Both sets of data are, in this way, directly comparable. In Col. 7 of the tables we indicate the confidence level of the variability, when observed; C is defined as $\sigma_{\rm T}/\sigma$, where $\sigma_{\rm T}$ is the standard deviation of the target differential lightcurve. The adopted variability criterion requires that for a variable source $C\geq2.576$. Finally, in Col. 8 we list the intranight variability amplitudes defined as (Heidt & Wagner 1996):

$$Y=\frac{100}{<D\gt}\sqrt{(D_{\rm max}-D_{\rm min})^2-2\sigma^2}\%,$$ (1)

where $D_{\rm max}$ and $D_{\rm min}$ are the maximum and the minimum of the differential lightcurve, respectively.

  

Table 2: Results for radio-loud AGNs

  

Table 3: Results for X-ray selected BL Lac objects

  

Table 4: Results for radio-quiet quasars

Among 15 radio-loud AGNs of our sample, 9 (60%) displayed microvariability with amplitudes from 2.2% up to $\sim\!8$%. Figures 2, 3, and 4 show lightcurves for two variable and one nonvariable radio-loud objects: the BL Lac 0537-441 (amplitudes $\sim\!\!7.7$% each night), and the flat-spectrum QSOs 1244-255 (with amplitudes of $\sim\!6.8$%) and 0637-752 (no variation detected). The short-term variations observed in 0537-441 are part of a larger outburst with timescales of $\sim\!1$ day and amplitudes of $\sim\!20$%. This can be clearly appreciated in Fig. 5, where we present the night-to-night behaviour.

  

Figure 2: Differential lightcurves for the radio-selected BL Lac object 0537-441 and for a comparison averaged star for the nights of December 21st and 22nd, 1997. The filled symbols represent (in all figures) the differences O -<C1i>, whereas the open ones are <C1i>-<C2i> (see Sect. 2 in the text). Error bars show the formal photometric errors (valid also for all figures)

  

Figure 3: Lightcurves for the RLQSO 1244-255 (filled symbols) and comparison (open symbols) obtained April 28th, 1998

  

Figure 4: Lightcurves for the nonvariable RLQSO 0637-752 (filled symbols) and comparison (open symbols) obtained December 22nd, 1997

  

Figure 5: Two-nights differential lightcurve for 0537-441

The objects in Table 3 are classified as XBLs. As noted by Heidt & Wagner (1998) on the basis of an extensive study of intraday variability, this type of BL Lacs seems to display different duty cycles and variability amplitudes than radio-selected sources. These differences seem to be present also at microvariability timescales: just 1 out of 3 objects in our sample (33%) showed microvariations. The variable XBL is 1101-232 and its lightcurve is shown in Fig. 6. The observed variability amplitude (3.6%) is also lower than the average amplitude of radio-selected AGNs (5.2%). If we restrict our set of radio-loud objects to radio-selected ones (i.e., those in Table 2), we get that 8 out of 12 (67%) showed microvariability. The average amplitude for RBLs is 5.12%, while for RLQSOs it is 4.36%. The fraction of variable radio-loud AGNs in our sample is lower than that determined by Heidt & Wagner (1996) for RBLs at intraday timescales ($\sim\!\!\!80$%). Although these differences are suggestive, the very limited number of objects in our samples makes necessary more observations in order to draw any statistically significant conclusion.

One of the XBLs included in our selection is the well-known BL Lac PKS 2155-304. Optical microvariability with timescales as short as one hour has been found in this object by Carini & Miller (1992) and Paltani et al. (1997). Heidt et al. (1997), however, did not detect any variation on timescales shorter than one day and have suggested that this object could pass through relatively frequent quiet stages at microvariability level. Our results, which span two consecutive nights without detecting any kind of variation, lend additonal support to this conjecture.

None of the 8 RQQSOs observed during our campaign have displayed variability strong enough to be classified as a variable source. Just in one case, the first night of observation of 2340-469, there seem to be some microfluctuations in the differential lightcurve at the level of $\sim\!3.3$%. However, the confidence for this variability is just C=1.7, so that the source formally classifies as NV. This QSO will be included in future monitorings in order to confirm whether it presents real variability or not. In Fig. 7 we show its lightcurve with average comparison star for the night of April 9th, 1997. Additionally, in Fig. 8, we show cross comparisons with different individual stars for that night. In no case can the object be classified as formally variable according to the adopted variability criterion.

  

Figure 6: Lightcurves for the XBL 1101-232 (filled symbols) and comparison (open symbols) obtained April 29th, 1998

  

Figure 7: Lightcurves for the RQQSO 2340-469 (filled symbols) and comparison (open symbols) obtained April 9th, 1997

  

Figure 8: Differential lightcurves for 2340-469 with respect to different stars in the CCD frame, and comparison curves formed from diffrences among the stars