The IRC name finally rang a bell: Bidelman's publication "Spectral classifications for stars of the Caltech Two-micron survey", Pubs. Warner & Swasey Obs, vol. 2, no. 6 (1980), which lists the M8 type as coming from Hansen & Blanco (1975 AJ 80,1011), which in turn derive from CTIO Curtis Schmidt plates. The type is almost certainly correct, although such stars are likely to vary not only in brightness but somewhat in spectral type. \Brian