I agree with Kato-san that this appears to be a pretty typical red variable. The IRAS 12/25 ratio is ~2, typical for semiregular or Mira. The odd USNO-A2.0 color almost certainly results from the red/blue sky survey plates having been taken at different times. (Only for POSS-I were the plates taken on the same night as a requirement.) There is no spectral type for the IRAS source in MacConnell's IR objective-prism survey, I think because it is a little too far off the galactic plane. \Brian