Following Z. Shunnar's publication of the Elephantine papyrus P. Berol 23 000 1970,1 the text was re-examined by J. Naveh and S. Shaked 1971,2 and R. Degen 1972.3 A number of superior readings have been proposed.4 A noteworthy omission all of these publications has been the lack of a systematic paleographic dating for the papyrus. Based on some tenuous name identifications, Shunnar dated the papyrus about 410 B.C.E. Both the publications of Degen and of Naveh and Shaked place the papyrus at the end of the fifth century or beginning of the fourth century. Based on a detailed analysis of the script of the papyrus, it can be shown that the papyrus dates the first half of the fourth century, preferably the second quarter of that century. Furthermore, there are one or two grammatical points that tend to support such a date. The importance of a precise dating, as opposed to a vague end of fifth century assignment, is that it provides us with new material a century that is sparsely represented epigraphically. Finally, a careful epigraphic analysis clears up certain disputed readings (e.g., mpntdt vs. spntdt, hby vs. sby, 'lmzbnwk vs. lmzbnh, etc.) and should lay any controversy to rest. For comparison with earlier material, Kraeling's papyrus no. 13, dated October 1,399 B.C.E. has been used and appears as number one the script table.5 This is a dated papyrus written a cursive hand and provides a good terminus ante quem for P. Berol 23 000, also written a cursive. The choice of a terminus post quem is complicated slightly by the fact that there is so little fourth century cursive material available. The two manuscripts chosen, the Louvre Papyrus6 of the mid-fourth century and two of the ed-Daliyeh papyri7 (352 B.C.E. and 335 B.C.E. respectively), while definitely belonging to the cursive tradition, are a more formal hand and, consequently, tend to be a little more conservative. 'lef: The 'dlefs this papyrus all have the dis inctive fourth century crescent-shaped leg. The letter is made three strokes, as can be seen by the (apparently) aberrant example 1. 5, lytm' (which cannot be a s as Shunnar read it). The first stroke made is the right arm. Then the diagonal stroke is made rom upper left to lower right. Finally, the left leg is made a crescent shape from left to right. In lytm', this stroke failed to connect with the diagonal second stroke. For a similar 'lef, cf. 1. 2, y'mr. Consequently, the 'alef of lytm' is not as unusual as it appears at first glance. In b'gr 1. 3, however, e is an unusual 'llef, where it appears that the crescent leg was made more as a straight line overlapping the diagonal. 'Alef is not a happy reading here, though it does not seem to be anything else. bet: Just as Kraeling 13, bit still has its round curve from the fifth century. The right shoulder has straightened out, however, and the base is elongated and rounder, as is normal for the fourth century. It is well on its way to becoming the bit found the Louvre Papyrus and ed-Daliyeh, where the head disappears and the curve becomes the pronounced feature of the letter. ddlet: The ddlet, form, is virtually indistinguishable from rs', though dailet tends to have a rounder shaft, at least the ideal form. It still preserves the right shoulder of the fifth century, which is practically gone the Louvre Papyrus and completely lost Diliyeh I. Shunnar's early reading 1. 7, byswn, in Syene, cannot stand. Although not particularly well made, the third letter is clearly a dalet (and just as clearly not a slmek), cf. 1. 3, ydh for a similar dalet. For kaf instead of Shunnar's final nan, see below. hJ: The hi is a most unusual letter, very advanced and one of the signs pointing to a clear fourth century date. The papyrus maintains a clear, consistent distinction between the medial and the final forms of the hJ. This is a feature which is never I Shunnar (1970) original publication and plates. 2Naveh and Shaked (1971) 379-82. 3 Degen (1972) 9-22. SOf late, Rudolf Macuch has come out defense of many of Shunnar's earlier readings. See Macuch (1973) 58-60. 5Kraeling (1953). 6 Cross (1955) 147-72 and fig. 2. 7Cf. Freedman and Greenfield (1969) pl. 36 (facing p. 120).