On a Correlation between the Radial Velocities of Optical and Radio Interstellar Lines.

Abstract

The radial velocities of the interstellar K line of Ca ii in a total of 39 high-latitude stars ( b ~ 150) having single interstellar line components have been analyzed in conjunction with the corresponding H I radial velocities of the 21-cm line profile. The correlation coefficient is found to be +0.65~*0.09 (m.e.). No radial velocity in the sample, either radio or optical, exceeds ~ 12 km/sec. When the H I radial velocity is subtracted from the corresponding optical value and a histogram of the points is plotted, the mean value of the difference (~ AV>) is +0.3t0.6 (m.e.) km/sec, showing that the large- scale motions of H I and Ca ii are statistically the same. The standard deviation of the AV histogram is 3.5*~0.4 (m.e.) km/sec. After this figure is corrected for instrumental effects, we find a cosmic dispersion of approximately 3 ~ 1 km/sec, which we attribute in varying degrees either to cloud rotation, to differential Ca-H streaming, or to turbulent motions on a scale that is smaller than the solid angle of the 20 radio beamwidth but large enough to include the optical solid angle through the interstellar cloud. The last alternative seems the most probable. From the area under the 21-cm line profile we can derive the number of H I atoms per cm 2in the line of sight (NH). Then, by adopting a hydrogen-tocalcium abundance ratio, we can estimate Nea (em-2). The optical lines are unsaturated with K/H ratio ~1.6. We can derive Ne ii (cm-2) from the usual curve-of-growth analysis given by Str6mgren (Stro~mgren, B., Astrophys. J. 108, 242,1948). The range of the ratio Nea/Nca ~~ over ten stars is 190/XY to 4500/XY, with a mean value of 1400/NY; the same ratio predicted theoretically is 83 Z/nu Pv, where N--the factor by which calcium is underabundant compared to the solar value (NH/NCa -6.5 X 10 N) Y the total H I optical depth/H I optical depth in front of the star; 7--the factor of overionization compared to Seaton's estimate, viz., flea iiifle =0.025Z cm-3 AT~a iii-Nea; flCa II ~V--the ratio of true-to-estimated electron density (n6-~3X10~4nu W cm-3). Most of the fluctuation in the observed ratio of N~a/N~a ii may be attributed to fluctuations in either the neutral hydrogen density 13~ or the ionization field 7 from star to star. However, on the average, the observed and predicted values disagree considerably in the sense that less Ca ii is observed than is predicted. An underabundance of calcium could account for this disagreement. A repetition of this analysis over five stars shows agreement between the observed and predicted NNa/NNa ratio. We note with interest that Spitzer and Routly (Spitzer, L., Jr., and Routly, P. M. Astrophys. I. 115, 227, 1952) have shown that calcium is underabundant relative to sodium in lowvelocity interstellar clouds.