Of the the Motions of the Magellanic Clouds

In 1917 I called attention to an apparent systematic difference between the radial velocities of the gaseous nebulae in the northern and southern parts of the Large Magellanic Cloud, suggesting that this difference might result from a rotation of the cloud. In 1920 Hertzsprung proposed the simpler explanation, namely, “that all parts of the cloud are moving parallel and with the same speed in space.” He determined the co-ordinates of the space motion from the radial velocities of the 17 nebulae in the Large Cloud, and showed that the velocity of the single nebula observed in the Small Cloud “is consistent with the supposition that the two clouds share the same motion.” This was before anything was known about the effects of galactic rotation on radial velocities. It is now clear that the rotation of the galaxy accounts for the major part of the velocities measured in both clouds. Yet, after corrections for galactic rotation (Table I, column 5) are applied, there remain residual velocities which vary roughly with galactic longitude (Table I, column 6). The exact value used for the constant of galactic rotation is not important. We have used 275 km/sec. Any reasonable increase or decrease of this value will produce a decrease or increase in the mean residual velocity, but will always leave a difference in the means for groups of smaller and of greater longitudes. This fact leads one to inquire whether or not the comparatively small residuals remaining after correction for galactic rotation suffice to determine the space motion of the Large Cloud or to give any evidence that the two clouds share the same motion. Table I lists the nebulae in the Magellanic Clouds for which radial velocities have been observed. Their galactic co-ordinates are in the second and third columns ; their observed radial veloci-