The most demanding part of the work was amber drilling the perforation. Tubular beads were drilled from both ends. The work could be spoiled if the drilled channels from both ends did not meet, or if the piece broke in the course of drilling. Button-shaped beads were drilled at an oblique angle in two places, so that the channels met, forming a V-shape. It happened that the drill accidentally penetrated right through to the front of the piece, or else the amber broke at the back between the two drilled channels. If drilling was unsuccessful, a second perforation could be made. Thus, many pendants and button-shaped beads have three or four drilled holes in the back. Vankina suggests that drilling was accomplished by means of a flint drill or awl, using quartz sand as an abrasive. The drilled channels are usually conical and correspond to the shape of the ends of flint borers. She suggests that in drilling and smoothing of the walls of the drilled channels also bone needles were used, particularly for the cylindrical drilled channels through beads and pendants. Mazurowski has likewise concluded, from microscopic study and experimental work, that flint drills were used, as well as drills of bone awls and trimmed long bones of birds – the latter permitting longer channels to be drilled.
The retouched surface of the piece was ground with sandstone, leaving fine striations, and subsequently polished to obtain a smooth, lustrous surface. Often only the front was polished, economising on labour, a practice characteristically observed on button-shaped beads, where the reverse side is often left in the ground state. Experiments by Mazurowski showed that the striations from grinding could be most rapidly obliterated by rubbing the object on material such as inner leather dusted with loam or chalk dust, a glossy surface being obtained by polishing on a piece of outer leather or woollen cloth.
Pendants of irregular shape were more simply made: flaking was minimal or entirely absent, largely retaining the natural form of the amber lump and giving the surface a partial polish.
Amber artefacts
In the analysis each amber artefact in the Sarnate collection, kept at the Archaeology Department of the Latvian History Museum was classified in terms of artefact group, type, degree of surface finish and state of the perforation. Based on the typological scheme of Vankina, the artefacts were divided into the following groups and types:
– button-shaped beads (circular, oval or rectangular in plan, circular or plane-convex in cross-section);
– tubular beads (cylindrical or barrel-shaped);
– pendants (trapezoidal, droplet-shaped/elongated, rounded, conical, other forms, irregular);
– discs (cross-section lenticular or plane-convex);
– rings;
– indeterminate, partially shaped pieces;
– flakes and split lumps;
– unworked amber lumps entirely covered in cortex.
The following degrees of surface finish were distinguished:
– natural cortex all over;
– partially covered in natural cortex, partially flaked;
– surface entirely flaked;
– partially flaked, partially ground;
– surface entirely ground;
– partially ground, partially polished;
– surface entirely polished;
– front polished, reverse ground (mainly button-shaped beads);
– partially natural cortex, partially polished (irregular pieces not shaped by flaking, only polished).
The following perforation states were distinguished: a unperforated (intact);
– partly perforated (intact);
– perforation complete (intact);
– broken at the perforation, with perforation incomplete;
– broken at the perforation, with perforation possibly or certainly complete;
– fragmentary piece not showing a perforation (such as the lower part of a pendant or the margin of a button-shaped bead).
In order to establish the manufacturing sequence more precisely, information was brought together on the surface finish of those pieces that had unfinished perforations. This should reveal to what extent the surface was worked before perforation. Out of four tubular beads with an unfinished perforation, two had a flaked surface, one had a partially ground surface and one was completely polished. Out of 1.7 button-shaped beads with an unfinished perforation, in one case (6%) the surface was only partially flaked, in four cases (24%) it was entirely flaked, in 10 cases (59%) it was entirely ground, in one case the front only was polished, and in one case the whole piece.