Simultaneous observations in the SiO J=1-0, v=1 and 2 transitions at 43.122 and 42.821 GHz, respectively, were made with the 45-m radio telescope at Nobeyama in the period from January 1997 to May 1998 (see Izumiura et al. 1999).
We have chosen a sample of sources in the galactic plane area, -10 <l<35 and |b| < 3, from the IRAS point source catalog. The source selection has been made in terms of the IRAS 12 flux density, the color, and the measurement quality index;
Jy,
,
,
and 
are the IRAS 12 and 25
flux densities of a source, 
and 
are the measurement quality indices at 12 and 25 , respectively,
and ``3'' means that the quality is high. These selection criteria
effectively extract dust-enshrouded objects
with 
K from the IRAS point source catalog.
Above criteria are quite similar to the criteria used in the
previous bulge SiO maser survey ([]; [Izumiura et al. 1995a];
[Izumiura et al. 1995b]), and the effectiveness of the criteria
for the SiO search has been well confirmed.
Near-infrared photometric observations of IRAS sources
(see Deguchi et al. 1998) were made
with the ANU (Australian National University) 2.3-m telescope at Siding
Spring Observatory, Australia, on 1997 June 18--23 and 1998 June 9--15
using the infrared array camera CASPIR
(the Cryogenic Array Spectrometer/Imager). The
camera uses a 256 
256
InSb detector array with 
per pixel, covering a field of view
of 
. The seeing size was generally about
--
.
The telescope pointing was accurate enough to locate the commanded
position (IRAS PSC position)
within an error of a few arcseconds.
To overcome bad pixels scattered around in the detector array, we
observed with position offsets of 
in declination for every
source.
Because most of the target sources were quite bright in the K-band, a
very short exposure time was used [0.3 s (H and K) and 1 s (J),
100 cycles].
Even with a short exposure time, an
appreciable number of sources were saturated in the K-band. In such
cases, we took an image with 
pixels and used it for the
photometry.