The question of where and by whom and how books were kept and made ready for users is not one that has been very much discussed although the questions who were the librarians and where were the books kept has been more or less implied in the discussions of temple versus secular collections. Mr. Tedder’s dictum that “these records would naturally be preserved in sacred places, and accordingly the earliest libraries of the world were probably temples and the earliest librarians priests” is modified and perhaps at the same time confirmed by the history of pre-phonetic libraries. It is true that in primitive tribes the medicine man is generally a keeper of records, but it is true also that among the Mexican Indians certainly, and pretty clearly among North American Indian tribes and in many African tribes, the shaman or medicine man is not the only keeper of records. It is true also that in the early Egyptian practice the priests were the keepers of the books whether it was in the temple, archives or the palace archives, but even here it seems to be the fact that there were military records, department records, and local administrative records in the different nomes kept by scribes who were not priests.

The keeping of records must in fact have begun before there was any special place, even the simplest hut or medicine wigwam or cave, set apart for distinctively religious purposes, although the setting apart of such places is apparently as old as the caves of the Stone Age. With these qualifications, the history of votive offerings tends to confirm the statement that the earliest public or tribal libraries were religious and the corresponding librarians the priests.

In very early times, and in much later times among primitive peoples, even the art of writing itself was often kept as a secret mystery in the custody of priests. The name “hieroglyphics” points in this same direction, and the temple collections of sacred books, the so-called books of Thoth and books of Hermes, point in the same direction. In general, however, this monopoly of letters seems rather to have been a deliberate assumption by the priests, as it is sometimes assumed by savage royalty, rather than the original situation. It applies, of course, rather to newly devised kinds of symbols, such as the vast number of systems of secret writing which have been evolved in all ages, than to the ordinary current record methods. That some of the earliest libraries were secret libraries, however, is an interesting fact, and one which may throw light on the mysterious collections of shrines and portable collections of objects in the liturgical processions in Egypt.

The methods used by these priest librarians for keeping and using the books form in themselves an interesting and little studied subject of very considerable extent.

The different kinds of writing required different sorts of receptacles. The book chest or bookcase, from which has come through the Greek the common word for library in languages other than English, was the most universal and natural way of keeping almost every kind of tangible record. The wooden chests and clay chests of the earliest historical periods must have extended well back into the pre-phonetic period and have also been found among primitive and semi-civilized peoples. They can obviously be used for quipus, message sticks, or almost any portable document. The same is true of the clay jar so often used in the earliest historical period. In the case of wandering tribes, however, less rigid or fragile materials are certainly better, and the book pouch was, therefore, in very early, and probably much earlier use than either boxes or jars. The skin pouch, like the skin water jar, is naturally suggested and easily made. This early form survives in the medicine bag, the lawyer’s green bag, and the schoolboy’s bag as well as in mail pouches for post-office use.

The use of basketry work and perhaps other textile work as bookcase also certainly extended back into pre-phonetic times and is represented in primitive usage.

It is not to be supposed, of course, that in these primitive times there were often separate buildings, such as the later Greek treasuries, or even separate rooms as in the Egyptian temples, and the archives at Boghaz Keuei and elsewhere, although separate huts for these and especially for “collections of liturgical objects” would perhaps be almost the first use for covered rooms, while sacrificing was still conducted in the open air.

Something like a pouch or wallet must have been used for the marked pebbles of the Stone Age and for pebble counting generally before the grooves and rods of the abacus were invented.

The method of keeping and displaying the books in the boxes, pouches, rooms or buildings, varied of course according to the nature of the document. In the modern library there is a great difference between the machinery necessary to keep and display folded documents, rolled documents, and ordinary bound books. The pouch may have had compartments like a modern purse. Basketry, clay and wood cases did have compartments, one for each roll, in quite early papyrus days.

In some of the late Babylonian libraries the clay tablets were evidently displayed on shelves but they were more commonly kept in clay boxes or jars, alabaster boxes, and the like, after the general fashion of the treasuries in earlier times and until the quantity became great. Twig records were tied together in bundles, and the stringing together of records was one of the earliest and most extensively used methods. It may perhaps be said that it was the typical method of the earliest records. It is found in the stringing together of trophy objects for wearing on the person—necklaces, girdles, and draped strings of various trophies. It is found also early in the history of the abacus where the perforated pebbles or beads were strung on different rods set in the ground, and it is of course found in the developed abacus. The perforations of tablets, bearing the year marks, among the objects from the earliest dynasties at Abydos, suggest a stringing together of these annual records, although it is of course possible that these are labels and the perforations used to attach them to boxes. The analogy with annual records of primitive people, however, suggests this stringing together.

What may be called classification of these libraries is found very early. It is reflected perhaps in the early distinction between temple and palace libraries, and more clearly in the primitive distinction between shamans and secular recorders. The putting of like kinds of works in boxes together, medical works, etc., is found as early as 2700 B.C. in Egypt and quite early in Crete. The labels of Crete point to a classification of objects if not of object records.

When collections are small no cataloguing is necessary excepting in the librarian’s mind, and his first mnemonic aid is classification, which is in fact a sort of cataloguing and takes the place of all other cataloguing. It is to be noted that in the very earliest records the librarian goes with the king or the investigating committee when they go to look up the records.