Page:Popular Science Monthly Volume 58.djvu/422

414 gone into numerous smaller stars. Accepting this view, it would follow that the material in question was a sheet so thin that the thickness of the space filled by the cluster was an appreciable fraction of that occupied

by the stars. In other words, one-fifth of the stars of the region would be contained in a thin sheet. This result seems too improbable to be accepted.

The other and more likely conclusion is that the number of very