Page:American Historical Review, Vol. 23.djvu/141

Rh

When a term used in one language can be shown to belong to another used by a different tribe, the fact is valuable not merely as indicating diffusion, but the direction of diffusion as well (p. 69).

Specific enumeration of the author's conclusions must stop at this point, but before closing one is tempted to emphasize the double significance of Dr. Sapir's contribution. On the one hand, it kindles the hope that the deficient historical perspective in ethnology will in time be offset, at least in a measure, by the rigor of reconstructive technique. On the other hand, the Study in Methods is symptomatic of the new spirit of ethnologic science, which, having gathered in vast stores of descriptive data, begins to take stock of its resources, and sets about the task of interpretation and reconstruction with a method progressively more critical and precise, and under the guidance of a rapidly maturing body of theoretic doctrine.

an interval of nine years since the publication of volumes I. and II., the third volume of de Sanctis's great history makes its appearance. In the first volume, a criticism of the tradition and a description of contemporary institutions were interwoven. In the second one, with the formation of the Latin League the narrative element comes into somewhat greater prominence, and when we reach the war with Pyrrhus, toward the end of this volume, a reasonably consecutive and trustworthy narrative is possible. In this last installment of his work the author has taken another step forward in his method of treating the subject, by giving us a continuous narrative in his successive chapters, and by consigning his treatment of critical questions to appendixes and foot-notes. In discussing in a brief review a volume which contains over 1100 pages and nearly 1200 foot-notes, we shall be obliged to limit ourselves to comments on the scope of the work, on the author's critical attitude, and his conclusions on two or three very fundamental questions.