Cats. 550–51, purchased in Syria, have in common a carinated oval body, large filling-hole surrounded by a ridge, and a conical stub handle. And yet they differ slightly: cat. 550 has an almond-shaped body, while the larger body of cat. 551 is merely ovoid. Cat. 551 has a nozzle channel, while cat. 550 has an almond-shaped ridge that extends onto the nozzle and around the wick-hole. On cat. 550 the base is almond-shaped, but cat. 551 has a raised circular ring making a foot. The position of the stub handle differs: on cat. 550 it is placed on the oval ridge, on cat. 551 on the lamp shoulder.
A close parallel to cat. 551 is Bailey BM III, Q 2328, pl. 60, of a type that he suggests developed in the early Islamic Ummayad period. As for cat. 550, its green vitreous glaze is typical of Islamic pottery. The difficulty in classifying these Islamic lamps, as Rosenthal and Sivan 1978 (p. 129) point out, comes from the interdependence of early Islamic pottery and lamp production with the late Byzantine ceramic tradition, out of which it developed. This is seen in the oval shape of the lamps as well as in the decorative repertory, including a channel on the nozzle; both conical and tongue handles are also anticipated in the Byzantine lamp tradition (Rosenthal and Sivan 1978, pp. 128–29). Excavations at Khirbat al-Mafjar near Jericho, Ramleh, Pella, Beth Shearim in Palestine, Antioch-on-the-Orontes, and Fustat, as well as the studies of Day and Bagatti, are relevant for dating Islamic lamps (Rosenthal and Sivan 1978, p. 130).
The authors conclude that “we can only arrive at a general date for these [Islamic] lamps, ranging from the sixth to the eleventh centuries A.D., . . . [but] the sixth-century date proposed by Waagé [Antioch-on-the-Orontes] is too early although the lamps flourished in the seventh-eighth centuries. . . . The longevity of the type can be seen at Fustat, where it is well attested until the tenth century A.D., but only sparsely in the eleventh century, at which time other lamp types had already made their appearance” (Rosenthal and Sivan 1978, p. 130). Other dating suggestions include Fellmann 1975: from the beginning of the fifth until the eighth century; and Bailey BM III: seventh and eighth centuries A.D.