We observed a total of 182 sources. We found 24 certain and
5 possible
OH masers. This represents a 12.6 percent detection rate. This value rises
to 15.4 if the 5 possible OH masers are included. For all
detected masers, the ratio 
is always less
than 2.5 percent which confirms the conclusion of Cohen et al. (1988) that
IR pumping of OH masers is plausible. The correlation of OH at 1665 MHz and
infrared flux at 60 
is shown in Fig. 3 (click here). All masers but one fall below the
line 
. In this survey we did not detect any
unquestionable masers of Ib type with dominant emission at 1667 MHz (the
definition of Ib type is given in Paper I). However four weak possible
candidates were detected.
Also five possible OH/IR stars with typical double - peaked
line profile at 1667 MHz were detected.
We detected thermal emission towards 51 sources, and 4 sources were
observed in absorption. OH emission is supposed to be thermal when the
1665/1667 line ratio is between 1 and the LTE value of 1.8, and no
polarization is present. In our sample, thermal emission sources are weaker
than the masers (the maximum flux of 0.64 Jy towards 16244-2432 in
cloud, in the rest of the sample flux does not exceed 0.4 Jy,
the mean
value being 0.19 Jy). Several sources from our list of thermal detections are
difficult to classify with confidence and must be studied separately to
determine whether they are masers or thermal emission sources.
03439+3233, 06587-0852, 16235-2416 and 19005-0445 may be
candidate OH
masers of type Ib because of their apparent dominant emission at 1667 MHz,
the 1667/1665 line ratio being greater than the LTE value of 1.8. Also,
05387-0924 may be a candidate OH maser of type Ia.
Figure 3: A correlation between OH and IRAS fluxes. 1665 MHz fluxes of the detected
masers are plotted against IRAS fluxes at 60 
. All masers but one fall below the
line 
The spectrum of 05487+0255 consists of two narrow components
separated in velocity by approximately 3 
. This is probably
evidence of rotation of a cloud where the line is formed.
06067+2138 shows a remarkable broad wing on the red side of the
spectrum which spreads from 5 to 20 
. Wouterloot &
Brand (1989)
detected CO emission in this source at the radial velocity shifted
from the maximum of the OH line by less than 1 
. The CO line
profile also shows a red asymmetric top. This is probable evidence for a
high - velocity molecular outflow both in CO and OH lines.
Finally, 36 among the 54 detected thermal sources are associated with
dark nebulae from the catalogue of Lynds (1962). The
estimates of OH
column density made for thermal detections vary from
to 
with a mean
value of 
. As the values of N(OH)
range from
to 
in dark clouds (Magnani et al. 1988),
our results could be considered as an evidence that the thermal OH line
detected towards IRAS point sources are formed in interstellar dark
clouds.
The question may arise whether the association of IRAS point sources with OH thermal emission is accidental (IRAS sources are only projected on the clouds where OH lines are forming), or real (IRAS sources are embedded in the clouds). We compared our list of thermal sources with the list of IRAS cores of Wood et al. (1994). They studied IRAS images of nearby dark clouds and identified 43 "IRAS clouds" containing 255 "IRAS cores". 15 IRAS point sources towards which we detected thermal OH lines are associated with IRAS cores of Wood et al. (1994). As the results published by Wood et al. (1994) are incomplete, this number might be greater. This means that a large fraction of the IRAS sources, towards which a thermal OH lines are detected, are really associated with the OH emitting clouds.