Fourteen years ago, in a book entitled "The Study of Stellar Evolution" (University of Chicago Press, 1908), I attempted to give in untechnical language an account of some modern methods of astrophysical research. This book is now out of print, and the rapid progress of science has left it completely out of date. As I have found no opportunity to prepare a new edition, or to write another book of similar purpose, I have adopted the simpler expedient of contributing occasional articles on recent developments to Scribner's Magazine, three of which are included in the present volume.