218

R. C. BOSANQTJET

of experience and manual dexterity which the ordinary purchaser was not
likely to possess. In terms of metal, the nodule may be compared to
unsmelted ore, the prepared block to the ingot of commerce. One such
prepared block, an irregulär prism, -09 high and -07 in diameter, is in the
Finlay collection, and is said to have been found in Melos.1

There was some exportation of unworked nödules. We found several,
the largest weighing over 1 lb. avoirdupois (about half a kilogramme), in the
Mycenaean town at Palaikastro in Eastern Crete. They seem to be pieces
which were too small or too awkwardly shaped for conversion into saleable
blocks, and may have been destined for some of the minor uses mentioned
in § 5.

In a Roman mosaic excavated by members of the British School on the
west side of the ancient town of Melos, obsidian was found to have beon
used with excellent effect for the black tesscllae. With this trifling exception,
the quarries remained unworked and forgotten for upwards of three thousand
years, until Fiedler rediscovered them in 1836 and for a moment conceived
the idea that they might be made to supply raw material for the manufacture
of glass bottles.

§ 2.—The Obsidian Workshop of the. First City.

The preliminary shaping at the quarries had for its object the produc-
tion of regularly formecl blocks, the chips Struck off being left on the spot as
refuse. In the next stage, illustrated by discoveries made within the town
of Phylakopi, the relation is reversed; the object is to obtain synimetrical
ribbon-like flakes, and the nucleus or core from which they have been
dctached is thrown away as worthless. These flakes have a very keen edge,
and are commonly called knives, or razors.

In 1896 the remains of a regulär factory of obsidian knives were dis-
covered in the Compartment B 5. 3, near the south-west angle of the town.
The deposit belonged to the First or pre-Mycenaean city, and passed under
a wall of the Second or Early Mycenaean city.'2 When I saw it at the close
of the season's excavations, it was a bed of refuse cores, used or refuse
flakes, and useless chips, and was upwards of '20 m. thick. The photograph
(Fig. 192), represents a selection of noteworthy speciniens which I made there
and then ; a series of normal types are figured on PI. XXXVIII. 19-28.

Taking the latter first, ]|) is a core or nucleus the nearer side of which retains tlie rough
outer coat of the original Qodule ; 20-22, cores from which a number of ilakes have been Struck ;
2.'i 27, flakes Struck from such cores, 27 being of morethan averuge length ; 28, a chip from an
carlier stage in the manufacture.

1 '.!.....pare Torquemada's description of the

blocks used in Mexico; they were pieces
"about cight inches long or rather more, and
as thick as one's leg or rather less, and
eylindrical." The llaking-tool was applied

"againat the edge of the front of the stone,
which also is cut smooth in that part"—a
phrase implying preliminary shaping. Quoted
by Tylor, Anahuac, p. 381.
- B.S.A. iii. 17 and iv. 24.