We see therefore repeatedly how mists with enclosed central stars originate. They remind us to a certain extent of the Galaxy with its clouds and stars and along the road thus suggested trials have been made to reach the solution of the riddle. The difficulty lies in the fact that the orbs whose collision create “new stars” are small, probably smaller than our Sun, while the mass gathered in the Milky Way most likely is trillions[2] of times greater than that of the Sun. It is true that we know a few unique stars, such as Arcturus, which exceed our Sun in size several tens of thousand times, but not even two such stars would account for the mass of the Galaxy, and furthermore the probability that two stellar bodies of such rare dimensions would collide is so very small that it must be left out of account. [2] American and French numeration; billions acc. to Swedish and English usage. Kapteyn’s star-drifts, containing many thousands or probably millions of stars, appear to furnish the bridge that leads to the solution of the riddle of the Milky Way. These drifts were once enormous gas-clouds, in mass probably several million times greater than that of the Sun. They also had an extension equal to trillions of stars. The probability for the meeting of two such gas-drifts is comparatively large and should not be much smaller than for the entrance of a star-drift into the Milky Way, an occurrence which actually has happened as shown by Kapteyn. When two such enormous gas-clouds meet, each with a cosmic velocity of about 20 km. per second, a long time would not elapse before the gas molecules in the region of interpenetration would be retarded in their original motions. An extraordinarily strong concentration and heating would occur in this territory, which is surrounded by the comparatively cold and heavy masses which remain unaffected because outside of the impact-area. A certain degree of equalization would naturally take place in the layers adjoining the boundary between active and inactive parts and the former would, furthermore, be set into a rapid spin around an axis perpendicular to the plane containing the two original motions. On account of the great viscosity of gases, particularly at high temperature, the central part would rotate as a coherent unit. Thus it would form a disk of gaseous matter. This disk would be thickest in the middle and would become thinner toward the edge where centrifugal force acts most powerfully. [Illustration: Fig. 3. N.G.C. 4594. Exposure = 2 hours. 1mm = 6”7. From _Proceedings of the National Academy of Sciences of U. S. A._]