Manuaba outfit. What can be said, incomplete though it naturally is, refers to the Manuaba performance; to similar cases in the same category with all reserve, to the Moon no details whatever, and even less to the great Heger I drums that followed very different techniques to answer very different requirements.
Figure 22: Another early moko from Alor.
The Manuaba mold must have been meant to be used several times in succession. The printing in itself brought no risks. Once the wax model had been built up and the decoration was to start, the mold parts that covered only one single quadrant of the all-over pattern had to be imprinted four times each. The wax plaque meant for the tympan (to be cast in combination with the cuff, apart from the mantle) likewise had to be imprinted four times in a four-fold repetition of the same design. In view of the many twistings and windings involved with the knob-and-braid pattern, the modellers may have preferred to have their designs incised in a piece of wood to be imprinted four times in succession (the Moon's tympan clearly shows the resulting seams).
When comparing a geometric design taken from a completed drum (Figure 11) with a similar one taken from the Manuaba mold (Figure 13) it is obvious that in the latter instance the f is shown reversed. Since in a Pejeng type drum all designs should stand out from the surface background (zero level), the designs on the mold had to be negative. The Moon's tympan displays a very complex pattern enlivened and clarified by means of a variety of "levels", in part hardly visible or tangible, in part with clearly seen lines and protruding knobs. The differences in height are, comparatively speaking, considerable, but they are always positive, above zero, and the mold to produce them had to be their exact reflection, negative. The grooves in the mold (wood or earthenware) must have been provisionally filled with wax before they were brought into contact with the all over wax layer.
The procedure followed by the former metal-workers at Manuaba may have been like this: first a clay core (for practical reasons, hollow) was shaped (in due course removed after the casting, to free the inside of the drum's body). This core or inner mold was covered with a comparatively thin layer of wax which ultimately had to be the exact prototype of the mantle's walls. The layer had to cover all sides of the mantle except for the top (tympan plus cuff) that was to be cast separately. The wax had to be prepared to be fluid during the smearing but to harden after being exposed to the air. The wax model had to be measured and tested for smoothness, thickness, etc. The best way to do this would have been using a standard printing mold which also could impress the decorative designs, by converting them from negative into positive. Apart from these designs certain three-dimensional parts (the handles) could be modeled separately by hand, in wax, to be fastened to the over-all wax cover. In the case of Heger I type drums that had to be modeled and cast to be all in one piece, this method had to be followed; in that of Pejeng type drums probably other tricks could be performed.
As a next step the entire wax prototype had to be enclosed in a clay crust or outer mold. Plugs to maintain the correct distance between the core and the outer mold during the casting, holes for letting fluid wax, gases, superfluous metal, etc. escape, had to be fixed. The crust was slightly heated, not more than needed for the fluid wax to escape (to be used again for the next drum). The molten metal could be poured into the extremely narrow spaces between the inner and outer molds, now used for the casting. The quality of the metal (bronze) had to be carefully considered to guarantee a smooth flow and in due course a satisfactory sound. Beforehand the bell-metal had been mixed and allowed to derive oxygen from the air, needed for some of the ores to function at all. The composition of alloys played an important role. There should be a comparatively high percentage of lead or tin in addition to copper to make it spread quickly through the narrow open spaces over such a very large area during the casting. One of the Pejeng type tympans, the one from Bebitra, contained the following ingredients: Copper 75.5%, Lead .09%, Tin 14.51%. But local circumstances played a part; different parts of a drum sometimes needed different alloys, and eventually the declining standards of commercial casters corrupted their products.
The lost wax procedure was eminently suited to the production of the more complex types of kettledrums, both in the Heger I class (using three-dimensional excrescences like frogs, handles and various grades of relief decorations at the same time) and in the Pejeng type category. Clay was cheap and everywhere available, the wax (a costly material at any time) could be removed and put aside for the next casting. And more particularly in the case of all-in-one-piece production: the material used was versatile and apt to accept any kind of unexpected forms. A complete Heger I drum as well as the mantle and the tympan-cum-cuff of the Pejeng type, each of them modeled and cast separately by using the lost wax method, could be delivered all in one piece.
But in the case of the Pejeng type drums there was the problem of how to fix that metal cap to the separate body and to make it into a workable unit if only by preventing that top getting detached. This practical aspect, indeed, must have been more urgent in the case of Pejeng metalworkers than in that of the Heger I casters who first of all cherished the ideal of visual and material "oneness": all-metal, all-in-one, there being no compromise in such matters.
The Pejeng metalworkers, consequently, could more easily permit themselves to use methods that in the eyes of their Heger type colleagues would have been unacceptable. As it happened, on account of the earlier relations with the one-sided membranophone tifa being part of the Pejeng type's prenatal history, they followed a procedure that at first sight seems a not very practical one, but suited the tifa connection better than anything else. Or so it seems. Instead of simply connecting the tympan-cum-cuff combination with the top of the separate body by telescoping the latter into the former (a method commonly used visually and materially in classical Indonesian stone architecture, demonstrated in Prambanan and other famous monuments), the Pejeng casters prepared a groove with a depth of (in the case of the Bebitra tympan) 1 cm in the slightly thickened lower part of the cuff, that corresponded with a raised border on the top of the underlying mantle. This arrangement, first noticed in the Bebitra instance, was later confirmed for the Pejeng Moon as well. This dowelled contraption that was to prevent the shifting of the cap required most scrupulous preparation and checking. The groove may have been cut in after the casting of the top (a risky business that could not easily be cured) or else the top of the mantle was cast first, including the ridge. After that the wax model of the cuff-cum-tympan was impressed unto its metal counterpart to make the two parts fit accurately. This latter method or some slightly different alternative would seem preferable. A later generation (represented by the Pacung drum fragments from northern Bali) used a different version of the same dowelled junction (Figure 14 right) that may have been easier to prepare.
The most striking element in the Pejeng Moon's silhouette is the wide protruding rim of the tympan, not less than 25 cm on all sides. This extreme salient is one of the main differences that separate the Moon from the Heger I family of metal drums, where such excrescences, and even smaller ones topping the bulging upper part of the mantle, are unknown. On the other hand, both Heger II and III (which do not appear in the present volume), lacking the bulging tendencies in their mantles, have a modest salient. The tympans illustrated in this chapter display distinct but still modest salients: 8 cm (vs. 64.9), 6 cm (vs. 55 cm) and 3.5 cm (vs. 51.5). Occasionally the protruding part makes a turn straight down over a few millimeters. The creator of the Moon went a considerable step further. He must have thought that a tall but comparatively heavy-set drum like this needed a suitable lining that ultimately hinted in the direction of the horizon and the wider skyline. He consequently elected this artistic final touch.
The "heads" (a striking element that made Nieuwenkamp and others speak of "the drum with the heads") have conspicuous round, staring eyes, pointed noses (damaged), and earlobes distended by disk-shaped ornaments. Similar heads, albeit highly simplified, are found on the Manuaba mold, and on certain apparently old mokos (to be discussed later). The other examples of the "classic" Pejeng type probably had similar decorations, but without further information we cannot be sure. There may be a connection between the heads and the decoration