that to hope for a complete elucidation of a problem so far removed from observation and experiment would be vain. The conditions of the mythopoeic age cannot be reconstituted; and were they to reappear through the destruction of civilizations, the reflective understanding would not be present to examine and record them. The difficulty which besets the problem of mythology, owing to the remote antiquity of the myth-making age, is to some extent removed by observing the operation of the mythopoeic faculty in the historic period. Given social circumstances similar, if even only in a limited degree, to those of the prehistoric age; given a defect of the critical and reflective faculty, an absence of fixed records, and a susceptible condition of the popular imagination, myths have always sprung up. While it is not, therefore, possible to find exact analogies to the conditions under which the Greek mythology originated, something may be gained by directing attention to mediæval romance. The legends which in Italy converted Virgil into a magician, the epic cycles of Charles the Great and Arthur, the Lives of the Saints, the fable of Tannhäuser and the Venusberg, the Spanish tale of Don Juan, and the German tale of Faust are essentially mythical. What is instructive about mediæval romance for the student of mythology in general is that here the mythopoeic imagination has been either dealing with dim recollections of past history, or else has been constructing for itself a story to express a doctrine. After excluding the hypothesis of conscious working to a prefixed end, we, therefore, find in these legends an illustration of the sense in which the symbolical and rationalistic theories can be said to be justified. In the case of Virgil, the poetry of Rome's greatest singer never ceased to be studied during the darkest years of the dark ages, and his name was familiar even to people who could not read his verse. He was known to have been a pagan, and at the same time possessed with what then seemed like superhuman knowledge. It followed that he must have been a wizard, and have gained his power and wisdom by compelling fiends. Having formed this notion of Virgil, the popular fancy ascribed to him all the vast works of architecture and engineering which remained at Rome and Naples, inventing the most curious stories to explain why he had made them. When we turn to the Carlovingian cycle, we discover that the great name of the Frankish emperor, the memory of his wars, and the fame of his generals have survived and been connected with the crusading enthusiasm which pervaded Europe at a later period. Border-warfare between France and Spain plays a prominent part in this epic, and gradually the figure of Roland usurps upon the more historically important personages. To "dig for a supposed basis of truth" in the Carlovingian cycle would be vain; yet the view is forced upon us that without some historical