It is a time-honored custom to begin a discussion of a subject by a brief sketch of its history, to outline what we used to believe about it, but have outgrown. My old teacher at Columbia, Professor Egleston, used to begin his lectures on metallurgy by explaining to us at great length the old practice, which even then was quite out of date, in order, as he said, that we should know what not to do. Most of us, I am afraid, rather begrudged the time so spent, as we wanted to get some grasp on present practice and probable future developments. Nevertheless there is more than a little advantage in knowing the history of any theory or practice, the trend of its development, the ideas that used to lie back of its methods; for it is almost always true that these discarded theories control our practice and methods . . .