1946/09/01 by N. U. Mayall · 1 citation
Physics and Astronomy · #Stellar, planetary, and galactic studies #Astronomy and Astrophysical Research #History and Developments in Astronomy
paper · doi:10.1086/144856
openalex publication_date 1946/09/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/04/04
Radial velocities determined with a low-dispersion (430 A/mm at H7) spectrograph attached to the 36-inch Crossley reflector are reported for fifty globular clusters. The list is substantially complete to ~ = 400 for the equipment used. A~l1 the velocities are for the integrated light of the clusters. At least four satisfactory spectrograms were obtained for each cluster; the residuals from the mean velocity for each cluster indicate an average internal probable error for a single plate of ± 22 km/sec. From an inde- pendent series of higher-dispersion spectra (130 A/mm at H-y) of fifteen clusters observed with the 36-inch refractor and from a comparison with velocities determined elsewhere, it is found that, in gen- eral, the systematic errors are small, only several km/sec. The average differences without regard to sign, however, indicate that the final velocities may be uncertain by 15-20 km/sec, especially for clusters ob- served only with low dispersion. Of the fifty velocities measured, nineteen are positive and thirty-one negative; all of them fall within the previously known range from +291 to -360 km/sec. The excess of negative velocities is due to selection, for the majority of the observed clusters are located in galactic longitudes (145°-325°), where the solar-motion component is negative. The spectral types, although difficult to estimate because of the composite and peculiar nature of the spectra, appear to range from A5 to G5, with an average of F7.6, but there is a preponderant number (thirty) included in the smaller intervaifrom F8 to G5. Comparison with Stebbins and Whitford's photo- electrically measured color classes shows little variation of color with spectral class, after the effect of space reddening is removed by the cosecant law previously deduced from the color observations. On the basis of a number of solar-motion solutions it is found that (1) no K-term is indicated; if one is introduced into the solution based on fifty clusters, K = -7.9 ± 18.8 km/sec; (2) there is no appre- ciable deviation of the computed apex from the one generally assumed at galactic longitude 55° and lati- tude 00; (3) the solar motion with respect to the clusters, V~, averaged from four different ways of treat- ing the 50 velocities, is V® = 175 ± 25 km/sec, but this value probably is too small because of ob- servational selection; (4) from separate solutions for the two groups of clusters in the hemispheres con- taining the apex and antapex and from a discussion of the degree of completeness in the two groups, it is concluded that the most probable value for the solar motion, with allowance for the clusters of unknown velocity, is V~ = 200 ± 25 km/sec. The mean residual motion without regard to sign is 101 km/sec. A cluster solar motion smaller by about 75 km/sec than hitherto obtained from half as much data suggests re-examination of the question of the sun's orbital velocity, V~. From a consideration of the most recent material for galactic objects, it is concluded that Vo = 280 ± 40 km/sec, i.e., unchanged from values heretofore generally adopted. The local group of nebulae, with certain qualifications because of the possibility of red-shifts and a component of the motion of the galaxy as a whole being included in their radial velocities, provide a check on V~. Several solar-motion solutions yield the round number Vo = 300 ± 25 km/sec, so that there now seems to be a real difference of Vo - Ve = 80-100 km/sec between the sun's motion with respect to the galactic center and to the system of globular clusters. The most obvious interpretation of the difference is that the cluster system partakes in the general galactic rotation and that it may, for dynamical reasons, be appreciably flattened. Attempts to find, in the peculiar velocities, evidence of a general or differential rotation of the cluster system have led to inconclusive results. The reasons are partly the inaccurate distances for a number of clusters, but especially the large dispersions in velocity and in distance from the galactic plane. Neverthe- less, the photometric distances for many clusters are still too imperfectly known to rule out the possi- bility of a small amount of flattening of the system for reasonable values of space absorption. The elliptical outline of the projected distribution of the clusters in the Andromeda nebula indicates that a globular- cluster system may be appreciably flattened. This circumstance suggests radial-velocity determinations for the clusters in that spiral as probably the most direct means for further investigation of the dynamics of a system of globular clusters. I. PREVIOuS WORK The only extensive series of observations hitherto made to determine the radial veloc- ities of the globular star clusters is represented by the pioneer work of V. M. Slipher,' * Contributions from the Lick Observatory, Ser. II, No. 15. The principal results of this paper were presented oraiiy at the dedication of the Mexican National Astrophysical Observatory at Tonanzintla, Puebla, Mexico, February 20, 1942. I Pop. Asir., 26, 8, 1918; 30, 11, 1922; 32, 622, 1924. A preliminary announcement was made in Pub. A.A.S., 3, 331, 1918. 29