Opening with an account of the census, the Book of Numbers evidently stood, from the first, quite distinct from the previous books as a composition or compilation. The mustering of the tribes gave an opportunity of passing from one group of documents to another, from one stage of the history to another. But the memoranda brought together in Numbers are of various character. Administrative, legislative, and historical sources are laid under contribution. The records have been arranged as far as possible in chronological order; and there are traces, as for instance in the second account of the striking of the rock by Moses, of a careful gathering up of materials not previously used, at least in the precise form they now have. The compilers collected and transcribed with the most reverent care, and did not venture in any case to reject. The historical notices are for some reason anything but consecutive, and the greater part of the time covered by the book is virtually passed over. On the other hand some passages repeat details in a way that has no parallel in the rest of the Mosaic books. The effect generally is that of a compilation made under difficulties by a scribe or scribes who were scrupulous to preserve everything relating to the great lawgiver and the dealings of God with Israel. Recent criticism is positive in its assertion that the book contains several strata of narrative; and there are certain passages, the accounts of Korah's revolt and of Dathan and Abiram, for instance, where without such a clew the history must seem not a little confused. In a sense this is disconcerting. The ordinary reader finds it difficult to understand why an inspired book should appear at any point incomplete or incoherent. The hostile critic again is ready to deny the credibility of the whole. But the honesty of the writing is proved by the very characteristics that make some statements hard to interpret and some of the records difficult to receive. The theory that a journal of the wanderings was kept by Moses or under his direction is quite untenable. Dismissing that, we fall back on the belief that contemporary records of some incidents, and traditions early committed to writing formed the basis of the book. The documents were undoubtedly ancient at the time of their final recension, whensoever and by whomsoever it was made.