Previous
abstract 
Next abstract
We have carefully searched the Atlas of Low-resolution Spectra (IRAS Science Team 1986, A\&A Suppl., 65, 607; Volk and Cohen 1989, AJ, 98, 931) for sources of the 13 $\mu$m emission feature associated with silicate emission at 10 $\mu$m first discussed by Little-Marenin and Little (1988, ApJ, 303, 305). We have identified 73 spectra for which the 13 $\mu$m feature is detected at 4$\sigma$ or better. In several cases, we have confirmed the existence of the feature using GLADYS, the Air Force long-slit 10 $\mu$m spectrometer, at the Wyoming Infrared Observatory.
Our sample of 13 $\mu$m emission sources are nearly all late M giants, the majority of type M6 or M7. Variability types exist for 54; over half are SRb variables, while the remainder are fairly evenly divided among Miras and classes SRa and Lb. Most of the sources have LRS characterizations of 14, 15, or 21-24, i.e. weak silicate emission at 10 $\mu$m. The shape of the silicate emission feature varies from a nearly classic 10 $\mu$m profile broadened at longer wavelengths to a double-humped profile with peaks at both 10 and 11 $\mu$m. In the scheme of Little-Marenin and Little, these shapes would be characterized as Sil+, Sil++, and 3-component. The root mean square radial velocity of our sample is 31 km/sec. The mean angle from the galactic plane is 29 degrees, and there are no obvious dependencies with galactic longitude. These properties indicate that our sample consists of old Population I AGB stars.
We have also investigated how the strength of the 13 $\mu$m emission varies with spectral type, class and period of variability, LRS characterization, 10 $\mu$m feature width, and galactic position. We find no correlation with any of these properties, indicating that the 13 $\mu$m emission is not unique to any specific class of Population I AGB stars. There is no strong evidence for the contention that the 13 $\mu$m sources are precursors to S stars.
