On the whole, the term “beginnings”, is better for the early periods than the term “prehistoric period”. “Beginnings” in this point of view differs from “prehistoric period” simply in overlapping a very little the shifting and uncertain borderland between the old prehistoric and historic, carrying over just far enough onto the firm land of annual chronological history to insure a safe footing in the field where written records begin to abound. In the case of books and libraries this line of division is most clearly made at the invention of phonetic writing, and this seems to correspond pretty well in time with the point of abundant written sources and of definite chronological data in the general history of mankind. In terms of relative chronology this line corresponds fairly with the first dynasty of Egypt. No doubt in its real beginnings it shades back far beyond its distinguishable first appearance at this time, but in broad terms it begins for Egyptians and Sumerians about this time, and even if this was not the earliest point of its appearance, it is the point at which the earliest abundant well dated and understood phonetic records are found. What time we shall count this to be in terms of annual chronology depends altogether by about 1000 years on whether we accept the views of the school of chronology illustrated by Breasted’s History or that for which Flinders Petrie is champion and in the same way with the Sumerian where King stands for the reduced chronology. When doctors disagree, prudent conservatism suggests the acceptance of that minimum amount on which both agree, in this case about 3400 years of the pre-Christian era. Without prejudice, therefore, to the possibility that Flinders Petrie may be right in putting the first dynasty a thousand years or so earlier, and remembering that even Breasted accepts a predynastic historic period extending to 4500 B.C. with a strictly historic period from “the earliest fixed date in the history of the world” in 4241 B.C., the division between phonetic records and earlier forms of written documents may be taken as falling at about 3400 B.C. At this time the invention of alphabetic writing was still perhaps two thousand years in the future but writing of some kind, mnemonic and picture writing, had already been practised for perhaps two thousand years or even much more. The beginnings, or the prehistoric, prephonetic and predynastic period of libraries, lie therefore back of the phonetic writing of 3400 B.C.—in picture book libraries, mnemonic libraries, object and memory libraries. § 11. _The evolution of record keeping_