Over the past few years several X-ray surveys have become available and many new BL Lacs have been discovered. Thanks to their strong X-ray emission, selection in this band has become the most effective way of discovering new BL Lacs. At present more than 50 per cent of the BL Lacs known have been discovered at X-ray frequencies.
2.3.1 The EMSS sample
The EMSS (Gioia et al. 1990; Stocke et al. 1991; Maccacaro et al. 1994) is a flux-limited sample of X-ray sources discovered serendipitously in 1435 Einstein IPC fields centred on high-latitude (|b| > 20°) targets. It covers 780 deg2 in the 0.3 - 3.5 keV band and goes down to fx ~ 5 x 10-14 erg cm-2 s-1, albeit in a much smaller area since the area of sky covered is a strong function of X-ray flux (see table 5 of Gioia et al. 1990).
The EMSS includes 34 BL Lacertae objects [plus two BL Lac candidates, which is
one less than in the original list, since
MS1332.6-2935, a BL Lac candidate in Stocke et
al. (1991), is now a confirmed
BL Lac (Perlman et
al. 1995b)], selected according to the following criteria:
observed equivalent width of any emission line < 5 Å and evidence
for dilution of starlight in the spectrum by a non-thermal continuum, which in
practice means a Ca II break with a relative flux depression blueward across
the break 
25 per cent in the
spectra (normal ellipticals would have values
around 50 per cent). Note that, although the equivalent width division is the
same as the one applied to the 1-Jy and S4 BL Lacs, here it refers to
the observed and not to the rest-frame value. It then
follows that, since
W
,rest =
W,obs / (1
+ z), the EMSS could classify as
AGN some objects that would have been considered as BL Lacs by the 1-Jy
sample, especially at high redshifts.
Browne & Marchã (1993) have suggested that the EMSS might misclassify some low-luminosity BL Lacs whose light is swamped by the host galaxy, which is typically a bright elliptical. Padovani & Giommi (1995a), within their hypothesis on the relationship between X-ray- and radio-selected BL Lac objects, have performed numerical simulations to establish the incompleteness level of the EMSS implied by the Browne & Marchã effect. By applying the prescription of Browne & Marchã to establish if a BL Lac is recognized as such or if it is misclassified, they found that about 10 per cent of the EMSS BL Lacs could be lost. Perlman et al. (1995a) have also discussed this effect and looked for possible misidentifications of BL Lacs with clusters of galaxies, coming up with one possible BL Lac (MS1019.0+5139) and four other (unlikely in their view) possibilities. We note that PKS 2316-423 (MS2316.3-4222), classified as a cluster of galaxies by Stocke et al. (1991), has recently been suggested to host a BL Lac object by Crawford & Fabian (1994).
Complete BL Lac subsamples from the EMSS have been presented by Morris et al. (1991) and Wolter et al. (1994).
2.3.2 The EXOSAT sample
The EXOSAT High Galactic Latitude Survey (HGLS: Giommi et al. 1991) is a flux-limited sample of X-ray sources discovered serendipitously in 443 Channel Multiplier Array (CMA) fields centred on high-latitude (|b| > 20°) targets. It covers 783 deg2 in the 0.05 - 2.0 keV band and goes down to fx ~ 2 x 10-13 erg cm-2 s-1, albeit in a much smaller area since the area of sky covered is a strong function of X-ray flux (see fig. 1 of Giommi et al. 1991).
The EXOSAT HGLS includes 12 BL Lacertae objects (plus two BL Lac candidates) selected according to criteria similar to those of Stocke et al. (1991).
2.3.3 The HEAO-1 sample
The HEAO-1 Large Area Sky Survey (LASS, also known as HEAO A-1) has produced a catalogue of bright, hard X-ray (0.8 - 20 keV) sources over the entire sky (Wood et al. 1984). The optical identification programme is still on-going and the final BL Lac sample has not been published yet. Schwartz et al. (1989) have presented some preliminary results, while Laurent-Muehleisen et al. (1993) have published a list of 29 HEAO-1 BL Lacs, one of which (2201+044) is now known to be a Seyfert 1 galaxy (Véron-Cetty & Véron 1993b). PKS 0521-365, previously discussed, has been included in our list as an uncertain BL Lac.
Although it is not entirely clear what were the precise criteria for the classification as BL Lacs, it is known that the HEAO-1 BL Lacs have been selected on the basis of their UV excess (Schwartz et al. 1989). This means that the HEAO-1 sample suffers from the same incompleteness as the PG sample, discussed above (Section 2.2.1).
Laurent-Muehleisen et al. (1993) give positions only for a subsample of their objects. This was not a problem for the majority of the remaining sources, since they are in common with other samples (Slew, 1-Jy). In one case (1H 0829+089) positions were derived from a search in radio catalogues around the IAU position (i.e. the position obtained from the source name) and therefore the coordinates are only good to within a few tens of arcseconds. In another case (1H 1914-194) no radio source was found within a degree from the IAU position, so no coordinates are available for this object.
2.3.4 The Slew survey sample
The IPC Slew survey has been constructed using the Einstein ``slew''
data taken when the satellite was moving from one target to the next (Elvis et
al. 1992), and covers a large fraction of the sky with sensitivities ~ 5
x 10-12 erg cm-2 s-1, reaching a flux limit
fx 
10-12 erg cm-2 s-1 over a much smaller
fraction of the sky (Schachter, Elvis & Szentgyorgyi 1993b).
Perlman et al. (1995b) (see also Schachter et al. 1993a) have presented a sample of 62 BL Lacs (which include 2 probable BL Lacs) extracted from the Slew survey adopting the same classification criteria as the EMSS survey. This is the largest BL Lac sample so far. Out of these objects, a complete sample of 48 BL Lacs has been defined. Six more objects have not been observed spectroscopically but have broad-band energy indices typical of BL Lacs and have been included here as BL Lac candidates. To these we also add 1ES1249+174W, discussed by Perlman et al. (1995b).
The HEAO-1 and the Slew surveys include many BL Lacs previously selected in radio surveys. This shows a severe limitation of a classification that has been frequently used in the recent past, based solely on the selection band. This classification divides the class of BL Lacs into X-ray-selected (or XBL) and Radio-selected (or RBL) depending on the band where the object was discovered: a number of objects could therefore be classified as both XBL and RBL. To overcome this difficulty Padovani & Giommi (1995a) have proposed to classify the objects on the basis of the ratio between the X-ray and radio fluxes (a parameter which univocally identifies an object) and suggested a division into ``RBL-like'' or low-energy cutoff BL Lacs (LBL) and ``XBL-like'' or high-energy cutoff BL Lacs (HBL). This division corresponds in fact to a break in their broad-band spectrum at infrared/optical frequencies for the former objects and at ultraviolet/X-ray energies for the latter.
2.3.5 The RASS sample
The ROSAT all-sky survey (RASS; Voges 1992) includes about 60000 X-ray sources, several hundred of which should be BL Lacs. The identification programme will inevitably take several years but early results have started to appear in the literature. Bade et al. (1994) report the discovery of 10 new BL Lacs through follow-up spectroscopy of AGN candidates detected in the RASS survey. Three of the objects are in common with the Slew survey. All optical spectra satisfy the EMSS BL Lac criteria of Stocke et al. (1991). Brinkmann et al. (1995), quoting a private communication from A. Kock, present five more new RASS BL Lacs which are (mostly) included in the 5-GHz survey of Condon, Broderick & Seielstad et al. (1989). Since no coordinates were given, we obtained them from the radio catalogue, searching near the coordinates obtained from the source name: they should therefore be accurate only to within a few tens of arcseconds. In one case (RXJ16264+3513), the radio source does not belong to the 5-GHz catalogue and therefore no coordinates are available.
2.3.6 The WGA and ROSATSRC catalogues
Two catalogues of ROSAT sources detected during pointed observations have recently become available: the WGA catalogue (White et al. 1994) and the ROSATSRC catalogue (Voges et al. 1994). These catalogues cover about 10 per cent of the sky with a much higher sensitivity than that of the RASS survey. Both catalogues include 50000 - 60000 X-ray sources and probably a few hundred BL Lacs. Identification of these sources has just begun. Wolter et al. (in preparation) report the discovery of one BL Lac and two BL Lac candidates. Giommi et al. (in preparation) also report the identification of a WGA source with a BL Lacertae object.