To the north it communicates with the Wadi Suweig along the steep and tortuous Wadi Dhaba, which is marked number 6 on the map. The word _dhaba_ signifies panther, probably in allusion to the feline animal which was at one time associated with the presiding female divinity at Serabit. Another valley, marked number 7 on the map, leads up to the temple ruins, with a path passing the mines which contain inscriptions in an early Semitic script. But it was not along these gorges, but along a path leading up from the plain El Markha along the Wadi Baba and the Wadi Nasb that the ancient Egyptians approached the sanctuary. This path has been included in the Ordnance Survey of 1871. A rock tablet marked P on the map, which commemorates the Pharaoh Amen-em-hat IV (XII 7), lies near the watershed, and this shows that the Egyptians passed here. A small Egyptian shrine (Q) was also discovered at the western end of the plateau. To the south of the plateau, and separated from it by several ravines and valley heads, extends the Wadi Umm Agraf which is comparatively remote from the valleys communicating with the Wadi Suweig and the Wadi Baba, and relatively close to the sanctuary. The approach to the temple from this side was unknown to Europeans till the winter of 1905-6, when Prof. Flinders Petrie and his party, who worked at excavating the temple ruins, pitched their tents here. A path was constructed from the camp up the mountain side to the temple by clearing away the stones. In parts an old path was re-used, the existence of which showed that there was at one time frequent intercourse between the sanctuary and the Wadi Umm Agraf. Some way down the Wadi Umm Agraf the valley floor is crossed by a wall made of rough stones piled together, the purpose of which was to mark off the upper reaches of the Wadi as is seen by a glance at the map. These upper reaches evidently constituted a tract of land the use of which was reserved to the sanctuary. A copious supply of good water is obtainable at a well some miles down the valley. The separation wall across the valley is undoubtedly old. A similar wall crosses the Wadi Maghara, which was dated by Prof. Petrie to at least the Fourth Dynasty. Its purpose, like that near Serabit, was to mark off the upper reaches of the valley, which in this case may have represented the _hima_ or tract of land that was originally appropriated to the sanctuary of the moon-god. [Illustration: VALLEYS OF SERABIT EL KHADEM Fig. 6—Sanctuary surroundings at Serabit. (Petrie: _Researches in Sinai_.)]