This section presents data collected from the literature on SiO, OH or H2O maser emission from S stars (relying on the compilation of Benson et al. 1990 for the earlier literature). These data are summarized in Table 3 (click here). The detection of SiO, OH or H2O maser emission is a clear indication that the circumstellar shell is oxygen-rich (see e.g. Deguchi et al. 1989, Lewis 1996), thus providing an important constraint on the chemical nature of the dust grains (see Sect. 4). In Table 3 (click here), "Y'' or "N'' in the SiO, OH or H2O columns means that the corresponding maser is or is not present, respectively, whereas a dash indicates that the maser has not been searched for in a given star.
The masers which are found in S stars are mainly SiO masers forming in the densest part of the circumstellar envelope near the stellar photosphere. OH maser emission has been detected in RS Cnc by Rudnitskij (1976) but not by Kolena & Pataki (1977), though the latter authors probed a different transition from that (1667 MHz) detected by Rudnitskij (1976).
The S stars with SiO maser emission
lie in the lower part of Region C, with some overlap with
Regions B and D (Fig. 12 (click here)).
As expected, the region occupied by the masers exactly matches the
region delineated by the stars showing the 9.7 
m silicate feature in
emission (IRAS LRS class E;
Fig. 5 (click here)). Indeed, 6 of the 8 SiO masers with an available
LRS class are of class E, the only exceptions being R Lyn and EP Vul (class F).
On the other hand, no maser emission is observed for stars
in Regions D and E (with the exception of EP Vul, but that star lies
very close to the boundary with Region C).
Figure 12:
Location of the maser sources in the (K - [12], [25] - [60]) diagram.
Positive detections
correspond to filled symbols, non-detections to open symbols, and
stars not probed for maser emission to crosses. Squares stand for SiO
masers, triangles for H2O masers and circles for OH masers