As regards the physical object, palæographers and philologists, who _restore_ to texts their original physiognomy, _restorers_ of pictures and of statues and other industrious toilers strive precisely to preserve or to restore to the physical object all its primitive energy. These efforts are certainly not always successful, or are not completely successful, for it is never or hardly ever possible to obtain a restoration complete in its smallest details. But the insurmountable is here only present accidentally and must not lead us to overlook the successes which actually are achieved. _Historical interpretation_ labours for its part to reintegrate in us the psychological conditions which have changed in the course of history. It revives the dead, completes the fragmentary, and enables us to see a work of art (a physical object) as its author saw it in the moment of production. A condition of this historical labour is tradition, with the help of which it is possible to collect the scattered rays and concentrate them in one focus. With the help of memory we surround the physical stimulus with all the facts among which it arose; and thus we enable it to act upon us as it acted upon him who produced it. Where the tradition is broken, interpretation is arrested; in this case, the products of the past remain silent for us. Thus the expressions contained in the Etruscan or Mexican inscriptions are unattainable; thus we still hear discussions among ethnographers as to whether certain products of the art of savages are pictures or writings; thus archæologists and prehistorians are not always able to establish with certainty whether the figures found on the pottery of a certain region, and on other instruments employed, are of a religious or profane nature. But the arrest of interpretation, as that of restoration, is never a definitely insurmountable barrier; and the daily discoveries of new historical sources and of new methods of better exploiting the old, which we may hope to see ever improving, link up again broken traditions. We do not wish to deny that erroneous historical interpretation sometimes produces what may be called _palimpsests,_ new expressions imposed upon the ancient, artistic fancies instead of historical reproductions. The so-called "fascination of the past" depends in part upon these expressions of ours, which we weave upon the historical. Thus has been discovered in Greek plastic art the calm and serene intuition of life of those peoples, who nevertheless felt the universal sorrow so poignantly; thus "the terror of the year 1000" has recently been discerned on the faces of the Byzantine saints, a terror which is a misunderstanding, or an artificial legend invented later by men of learning. But _historical criticism_ tends precisely to circumscribe fancies and to establish exactly the point of view from which we must look.