Such a position as this, assumed by one of the most minute searchers among modern historians, is highly interesting as illustrative of a reactionary tendency which will probably characterise the historical work of the near future. Hair-splitting analysis having been carried to its limits of refinement, there will probably come a reaction in the direction of a more comprehensive study of historical events in their wider relations. The work of the specialist, after all, is really important only when it furnishes material for wider generalisations. All minute workers in the fields of biology, geology, and the allied sciences, in the first half of the nineteenth century were unconsciously gathering material which, interesting in itself, became of real importance chiefly in so far as it ultimately aided in elucidating the great generalisation of Darwin. Perhaps the minute historians of to-day are in similar position. The special worker, imbued with enthusiasm for his subject, is apt to forget the real insignificance of his labours. Entire epochs are dominated by the idea of microscopic research, and the workers even come to suppose that microscopic analysis is in itself an end; whereas, rightly considered, it is only the means to an end. We are just passing through such an epoch as regards historical investigation. But, as just suggested, it seems probable that we are approaching a new epoch when the work of the specialist will be subordinated to its true purpose, while at the same time proving its real value as a means to the proper end of historical studies--the comprehension of the world-historical relations of events. CHAPTER II MATERIALS FOR THE WRITING OF HISTORY It is obvious that the materials for the writing of history consist for the most part of written records. It is true that all manner of monuments, including the ruins of buried cities, remains of ancient walls and highways, and all other traces of a former civilisation, must be allotted their share as records to guide the investigator in his attempt to reconstruct past conditions. But for anything like a definite presentation of the events of by-gone days, it is absolutely essential, as Sir George Cornewall Lewis pointed out in great detail, to have access to contemporary written records, either at first hand, or through the medium of copyists, in case the original records themselves have been destroyed. Lewis reached the conclusion, as the result of his exhaustive examination of the credibility of early Roman history, that a tradition of a past event is hardly transmitted orally from generation to generation with anything like accuracy of detail for more than a century.