From a purely theoretical discussion of the evolution of a mass of rotating gas we have been led to the hypothesis that the spiral nebulae are merely masses of rotating gas which have reached a stage of disintegration, the rotation having become so great through shrinkage that configurations of equilibrium are no longer possible. It would be of the utmost interest to follow out dynamically the different processes of this disintegration but unfortunately the mathematical difficulties have so far proved to be too great.

We have, however, found that the masses of these spirals must be supposed to be enormously greater than that of our sun, and the general nature of the disintegration has been seen to consist of the formation in the nebular arms of condensing nuclei each of mass just about comparable with that of our sun. Thus the hypothesis which has already been adopted seems to lead irresistibly to the conclusion that the final result of the process of disintegration which we see going on in the spiral nebulae must be the formation of star-clusters.

As to the features to be expected in these final star-clusters our dynamical analysis has so far told us almost nothing. It seems not unreasonable to expect that the star-clusters will be of the type we have described as “globular” – thus we may conjecture that the observed spiral nebulae are forming star-clusters similar to observed globular star-clusters and that the observed globular clusters have originated out of spiral nebulae.