Appendix I
The Dema Wall Saltcellar

During the excavation of the Dema house, a nearly complete black-glazed saltcellar was discovered "built into the rubble fill of the Dema wall" ("Dema House," 100, no. 99). Through the kindness of Professor J. E. Jones and Dr. B. Petrakos, Ephor of Antiquities of Attica, in 1978 I was able to examine this saltcellar, which was being stored with the finds from the Dema house in the Areos Street apotheke of the Greek Archaeological Service in Athens. Professor Jones also described how the piece was spotted by Mr. Sackett as he was sitting atop the wall: it was wedged between stones of the fill of the wall and had to be pried out with some difficulty. The wall-section from which the saltcellar came, between the railroad and sally port 11, had been reduced in height, evidently by stone-robbing (see DEMA 161). As a consequence, the original fill of the wall was more exposed here than is usually the case. Originally, the saltcellar was buried well within the fill of the wall. The piece is therefore securely associated with the construction of the wall, and its date provides a terminus post quem for the construction.

Concave-walled saltcellar Figure 43

Dema House no. 99. Single fragment approximately three-quarters complete. H. 0.027 m, D. of rim 0.059 m, D. of foot 0.062 m. Flaring ring foot, underside of foot rounded. Underside of bowl convex. Outside of foot continuous with curve of concave wall; curve of wall nearly symmetrical, extending slightly further below. Point of minimum diameter at midpoint of wall. Rim rounded. Floor of bowl rises in a continuous curve, inward turning at junction with rim. Good black glaze on all surfaces. Attic.

Published as catalog no. 99 in "Dema House," 100 (note that the diameter given there as 0.081 is an error for 0.061) and p. 101, with figure 7, p. 92, and plate 29.

The date of this piece given in the publication of the Dema house is mid fourth century on the basis of comparanda from the Athenian Agora, and it is therefore regarded as substantive confirmation of the 337-336 B.C. date for the Dema wall given in DEMA 186-89. There are a number of such saltcellars from dated deposits in the Agora, and they range in date from the late fifth century to the beginning of the third century. A series of examples arranged according to their deposit dates reveals a corresponding shape development according to which a date may be given to the Dema wall saltcellar. Such a comparison shows that the Dema wall saltcellar cannot be as late as the middle of the fourth century, and a date for the wall a good deal earlier in the fourth century than that proposed by Jones, Sackett, and Eliot is entirely possible on the basis of this, the only closely datable artifact which has a direct bearing on the construction date of the Dema wall.

Within the series of Attic concave-walled saltcellars of this type, there are a number of variable features, such as details in the profiles of the resting surfaces and rims and variable height/diameter ratios, which seem to have no chronological significance. The one feature that does change in a consistent manner over time is the profile of their concave walls. Examples early in the series have symmetrically curving walls, with the point of minimum diameter no higher than midway up the wall. By the second quarter of the fourth century, the curve of the wall becomes noticeably sharper toward the top, and the point of minimum diameter has correspondingly moved to about two-thirds of the way up the wall. These traits continue to develop into the third quarter of the fourth century, with the difference between the shallow curve of the lower two-thirds of the wall and the sharp curve of the upper third becoming more pronounced and producing a rim that overhangs the wall more noticeably than do examples from the first half of the century (cf. the description in Agora XII, 137).

This shape development can be observed in the series of saltcellars from the Athenian Agora illustrated in figure 43. The pieces are arranged according to deposit dates, from left to right beginning with the middle row:

P17422, published by Young, Hesperia 20 (1951): 195 and note 178, with plate 66c, as part of a group of sherds which "probably belonged to [a] table service . . . in the late fifth and early fourth centuries." Cited by Sparkes and Talcott as a parallel to no. 935, dated 425-400 B.C. From deposit A 20:6b, dated ca. 425-400 B.C., see Agora XII, 383.

Uninventoried, unpublished. From deposit Q 15:2, dated ca. 420-400 B.C.; see Agora XII, 397-98; Thompson, Hesperia 24 (1955):69-70; Crosby, Hesperia 24 (1955): 76-84.

P16951, unpublished. From deposit A-B 21-22:1, dated ca. 420-390 B.C.; see Thompson, Hesperia 16 (1947):210-11, and Agora XII, 384.

P27336, unpublished. From deposit I 13:2, published by Shear, Hesperia 39 (1970):212-19, with notes 101, 104. The deposit is part of a fill, probably a household dump, consisting of pottery which was "made and accumulated during the later years of the fifth century B.C. and the early years of the fourth, with a few pieces going as far back as the mid fifth century, and a few, more significantly, dating as late as ca . 370 B.C. " (Shear 219). Cf. Agora XII, 393, where the deposit is dated to the early fourth century.

P20, published as Agora XII, no. 936, dated 375-350 B.C. From deposit H 6:1, dated ca. 375-350 B.C.; see Agora XII, 392.

P12397, published by Thompson, Hesperia Supplement IV, 133, with figure 98d, as part of deposit G 12:23, dated to the second quarter of the fourth century, pp. 132-34. See also Agora IV, 239, and Agora XII, 391, for deposit G 12:23. This piece is cited in Agora XII as a parallel to no. 936, dated 375-350 B.C.

P12821, published as Agora XII, no. 937, dated 350-325 B.C. From deposit O 18:2, dated ca. 350-320 B.C.; see Agora XII, 396; Agora IV, 243; Thompson, Hesperia 23 (1954): 72-87.

The parallel for the Dema wall saltcellar cited in the Dema house publication is P12397 above, which has its point of minimum diameter well above the midpoint of the wall, with a sharper curve above this point than below. As can be seen in figure 43, the Dema saltcellar has closer affinities with the saltcellars in the upper row of the Agora group, where the saltcellars have symmetrically curving walls with their minimum diameters at the midpoint. In particular, the Dema saltcellar closely resembles P27336, both in dimensions (P27336: H. 0.03 m, D. of rim 0.062 m, D. of foot 0.065 m) and in profile, with its nearly symmetrical curve, extended slightly further below, and with the point of minimum diameter in the center of the wall.

On the basis of this comparison, the Dema wall saltcellar can be dated no later than the early fourth century. More precisely, it might have been in use any time from the last quarter of the fifth century through the first quarter of the fourth. As a terminus post quem for the date of the Dema wall, this piece indicates that the wall was built no earlier than the last quarter of the fifth century at the very earliest, and most likely after the beginning of the fourth century.