THE LUMINOSITY FUNCTION AT THE END OF THE MAIN-SEQUENCE - RESULTS OF A DEEP, LARGE-AREA, CCD SURVEY FOR COOL DWARFS

The luminosity function at the end of the main sequence is determined from V, R, and I data taken by the CCD/Transit Instrument, a dedicated telescope surveying an 8.25' wide strip of sky centered at delta = +2 8 degrees, thus sampling Galactic latitudes of +90 degrees down to -35 degrees. A selection of 133 objects chosen via R-I and V-I colors has been observed spectroscopically at the 4.5 m Multiple Mirror Telescop e to assess contributions by giants and subdwarfs and to verify that t he reddest targets are objects of extremely late spectral class. Eight een dwarfs of type M6 or later have been discovered, with the latest b eing of type M8.5. Data used for the determination of the luminosity f unction cover 27.3 deg(2) down to a completeness limit of R = 19.0. Th is luminosity function, computed a V, I, and bolometric magnitudes, sh ows an increase at the lowest luminosities, corresponding to spectral types later than M6-an effect suggested in earlier work by Reid & Gilm ore and Leggett & Hawkins. When the luminosity function is segregated into north Galactic and south Galactic portions, it is found that the upturn at faint magnitudes exists only in the southern sample. In fact , no dwarfs with M(I) greater than or equal to 12.0 are found within t he limiting volume of the 19.4 deg(2) northern sample, in stark contra st to the smaller 7.9 deg(2) area at southerly latitudes where seven s uch dwarfs are found. This fact, combined with the fact that the Sun i s located similar to 10-40 pc north of the midplane, suggests that the latest dwarfs are part of a young population with a scale height much smaller than the 350 pc value generally adopted for other M dwarfs. T hese objects comprise a young population either because the lower meta llicities prevalent at earlier epochs inhibited the formation of late M dwarfs or because the older counterparts of this population have coo led beyond current detection limits. The latter scenario would hold if these late-type M dwarfs are substellar. The luminosity function data together with an empirical derivation of the mass-luminosity relation (from Henry and McCarthy) are used to compute a mass function indepen dent of theory. This mass function increases toward the end of the mai n sequence, but the observed density of M dwarfs is still insufficient to account for the missing mass. If the increases seen in the luminos ity and mass functions are indicative of a large, unseen, substellar p opulation, brown dwarfs may yet add significantly to the mass of the G alaxy.