All this, however, is mere conjecture, while as an argument for the native origin of Egyptian glass we have the indisputable fact that the manufacture was carried on in the new town established by Akhenaten at Tell-el-Amarna (_circa_ 1450-1400 B.C.). This is made clear by the discoveries of Dr. Petrie in the winter of 1891-92. Among the waste-heaps of some important glass factories he has found enough material to put it beyond doubt that glass was there prepared from its raw constituents. First, with regard to the frits, the essential preliminary stage in the manufacture of glass: as I have said, some such half-fused material must have been long in use by the Egyptians in the preparation of their blue glazes. Complete freedom from iron was attained in this case (just as in after days by the Venetians) by the employment of crushed pebbles of white quartz as the source of the silica. These pebbles served also for the floor of the furnace, and they were doubtless more easily crushed after being thus used for some time. The fritting-pans, to judge from some large fragments of frit that turned up, were shallow bowls some ten inches across. These pans were, it would seem, supported for firing by cylindrical jars resembling the seggars of porcelain works. The shape and size of the crucibles in which the frit was subsequently melted may be inferred from some masses of glass found in the rubbish. These masses had been allowed to cool in the melting-pot, and the presence of frothy and worthless matter at the top was a proof that the glass was not merely remelted in them, but prepared on the spot from the above-mentioned frit. The glass was left to solidify in the crucible, and when cold, the crucible, as well as the scum at the top, was chipped away, leaving a clear lump of good glass. Dr. Petrie thinks that this glass was not remelted as a whole for subsequent working, but that lumps of suitable size were chipped off, and these, being heated to softness, ‘were then laid on a flat surface and rolled by a bar worked diagonally across them; ... the marks of this diagonal rolling are seen on the finished rolls.’ The rods thus produced were now drawn out to form a cane, or, if previously rolled flat, a thin ribbon. Beads were formed by winding these canes or threads of glass round a wire, or rather round a fine rod of hammered bronze, for wire-drawing was an invention of a much later date; such rods have indeed been found with the unfinished beads still on them. Similar canes of glass were doubtless worked in to the sides of the little vases to form the banded and chevron decoration which I have already described.