A: Athina, Greece, B: Athina, Greece
While information has survived concerning ancient Greek library and archive buildings from excavations of ruins, most information concerning library and archive holdings, and library and archive operation, is based on third party accounts, or is fragmentary or speculative. Dramatic exceptions to this overall lack of surviving archives from ancient Greece are the Archives of the Athenian Cavalry from the fourth and and third centuries BCE preserved on lead tablets. An Archive of the Athenian Cavalry was excavated in 1965 from a water well within the courtyard of the Dipylon, the double-gate leading into the city of Athens from the north. It included 574 lead tablets from the third century BCE. Six years later, in 1971, another hundred or so lead tablets from the fourth and third centuries BCE were excavated from a well at the edge of the excavated section of the Agora in Athens.
Historian of ancient archives Ernest Posner characterized these finds as
"by far the largest name file of ancient times. Tightly rolled or folded up, they contain the following information: the name in the genitive of the owner of a horse; the horse's color and brand, if any; and its value stated in drachmas, with 1,200 drachmas as the highest valuation given. Normally, only the name of the owner appears on the outside; the other data is relegated to the interior of the tablet and could not be read unless the tablet was unrolled or unfolded. A number of tablets are palimpsests; that is, the original entries were erased and replaced by new data" (Posner, "The Athenian Cavalry Archives of the Fourth and Third Centures B.C.", The American Archivist (1974) 579-82).
The wide range of pottery as well as lead tablets excavated from the Dipylon were described by Karin Braun in "Der Dipylon-Brunne B¹ Die Funde," Mitteilungen des Deutschen archäologischen Instituts Athenische Abteilung, Band 85 (1970) 129-269, plates 53-93. Plates 83-93 illustrate lead tablets unfolded to show the writing and tablets rolled up.
From the extensive information available, John H. Kroll, author of the primary paper on the 1971 excavation, developed a theory of the purposes and operation of the Athenian Cavalry Archives, of which I quote a portion:
"The continual turnover of the horses explains, I think, why the records of the horses' values were kept as they were-individually on lead tablets. Official annual records at Athens were normally kept in list form on papyrus or whitened boards. But since a cavalryman was likely to have changed his horse at any time in the course of a year, a more flexible system of records was called for-the equivalent of the modern card-file system-whereby the record of a given horse could be pulled out and replaced if the horse itself was replaced. For such individual records, lead had obvious advantages over paper or wood, and, becatuse it was cheap and could be erased and re-used repeatedly, it would have been less costly in the long run. The re-use of the tablets, incidently, must surely be a factor in the low survival rate of tablets in most series and the loss of other entire series. There is one other respect in which the tablets stand apart from most annual records. I assume that they were rolled or folded simply to facilitate storage and not because the evaluations they contain were to be kept secret. But the fact that they were folded or rolled up, many of them as tightly as they could be, indicates that no one expected them to be referred to on a regular basis. Indeed, since all of the unbroken tablets were recovered from the Kerameikos and Agora wells in their original folded or rolled state, it appears doubtful that any of the extant tablets had ever been consulted. This of course does not mean that the evaluations were never consulted, merely that the records were made up annually and filed away to be consulted only in rare, though anticipated, cases. If the occasion did not arise in the course of the year, they expired, were replaced with the next year's evaluations, and were put aside, eventually to be erased and re-used" (Kroll, "An Archive of the Athenian Cavalry," Hesperia XLVI  No. 2, 94-95). Kroll's extensive article occupies pp. 83-140 of the journal issue and includes numerous drawings and photographs.