Mccormick Spectral Statistics.

A summary is given of the spectral statistics of the second proper-motion program of this observatory. The 441 regions are fairly uniformly distributed over two-thirds of the sky. The pres~nt spectral counts are based on oniy 7600 classified stars, in contrast with the much larger numbers involved in the published spectral statistics of Harvard, Pots- dam, and Bergedorf. However, the McCormick material has the advantages that it samples two-thirds of the sky with a single instrument and that the number of samples is relatively large. Thus, although the total area classified is oniy four-tenths of the total area in the published portion of the Bergedorf Spektral-Durckmusterung,1 the number of samples is ten times as large. The spectra were obtained for use in connection with the second general proper-mo- tion program now being prepared for publication at this observatory. The program con- tains 441 regions, each covering 0.46 square degree, which are distributed fairly uniform- ly from the north pole to - 20° in declination. The regions were selected without any regard to apparent features of the Milky Way, such as star clouds, absorbing matter, etc. All the spectra were classified by the senior author from plates taken with the 10-inch Cooke prismatic camera, the dispersion being about 300 A/mm at Hy. The observa- tional technique, the criteria used for the classification of spectra, and the relation of the McCormick system of classification to other systems have been described elsewhere.2 The limiting magnitude of these spectral plates averages about 12m0 photographic. Inasmuch as the proper motions were determined with a photovisual telescope, the use of photovisual magnitudes is necessary. The system of the magnitudes has been carefully investigated by E. R. Dyer,3 who found it very close to the international photovisual system in scale, zero point, and color, so that any systematic error arising from a combination of these three causes is probably not greater than Omi. This is con- firmed by the satisfactory agreement between the McCormick spectral statistics and those of Bergedorf, as is shown in Figure 1 of the next paper. The disadvantage of using photovisual magnitudes in presenting spectral statistics is that incompleteness sets in among the K and M stars at much brighter magnitudes than among the early-type stars. Thus, there were many early-type stars which could have been classified on the spectral plates but which were not included because they were too feeble to be measured on the proper-motion plates; conversely, late-type stars of photovisual magnitude 11 were some- times too faint to be classified on the spectral plates. Since the stars measured for proper motion are not complete beyond photovisual mag- nitude 1 1'~'5, the counts presented in the left-hand side of Table 1 have been limited ac- cordingly. It will be seen that the proportion of unclassified stars is rather large in the low-latitude zones of the last magnitude group. This is caused chiefly by the over- lapping of spectra. Since we deal with photovisual magnitudes, it is easily seen that this overlapping would affect the late spectral classes much more than the early ones. This is quite obvious in the case of a partial overlap between an A and a K star, both of 1 Bergedorf, 1935 and 1938. 2 Vyssotsky, Ap. J., 93, 425, 1941