grammars; not only, for example, the Swedish-into-English exercises, but those from English into Swedish. The excellent Balbus may have begun Greek at seventy, but among moderns such a display of energy and industry at middle age is sufficiently remarkable. It should not be forgotten that Sumner, as his maiden publication witnesses, was a good Hebrew scholar, and that he knew Greek and Latin well. So that his control of languages, though he used to say that he was not quick at learning them, extended to some thirteen or fourteen; and of these he had an exact and precise grammatical control. It may be added that at about the same time he was acquiring a knowledge of calculus in order to see for himself what there was in mathematical economics. And all this while writing, lecturing, teaching a heavy schedule, and taking a leading part in faculty labors.
One of the characteristics of Sumner's mode comes out quite unmistakably in his essays; and that is his simplicity and clearness. He struck straight at the heart of a matter. He used to say that there were three questions to be asked about any production: What is it? How do you know it? What of it? Upon the last inquiry he laid particular emphasis; but, granted that there was any use in doing a piece of work, he was keen about his other two criteria: that it should be set forth so it could be understood, — that one should tell, with brevity and clearness, what it was that he had found, — and that he should give good and sufficient reason for his opinions. He used to prune the theses written under him of verbiage and slash out inexact expressions, usually making careful emendations, until the pages were scarcely recognizable. For himself, he abjured latinity and chose the tersest and most rugged of Anglo-Saxon terms, using, for an extreme example, a word like leechcraft in place of a more indirect and ponderous term. He hated long and involved sentences, and urged us all to be sure to translate German passages that looked as if they were significant, to see if they really were; for, as he said, "the German language and style lend themselves easily to bathos." He believed that if the thought were clear the expression would be, and
