Despite these practices, current research has confirmed that Marino Rivero's bandoneon was made by the Alfred Arnold factory, leaving it probably in 1930 as model 31a II (fig. 8), pre-assembled and tuned to a" = 880 the so called "orquesta tipica" pitch, which was only for export. In German-speaking countries, bandoneons were tuned to the "Normal A" (a" = 870) until 1941, when Germany joined the International Pitch Convention, at which time instrument makers switched to the higher a' for the domestic market as well.
Marino Rivero's bandoneon has been preserved in its original condition, meaning that all important components and their assembly remain as they were when the instrument left the Carlsfelder factory. It is the reed board, consisting of tongues mounted on long plates, that fundamentally defines the sonority of each bandoneon. As a result, the musical instrument manufacturers of better reputation had long been making their reed boards in their own workshops, employing their own well-guarded materials, tools and techniques. Only when the National Socialist government took over the direction of the German economy, and particularly after the outbreak of war in 1939, when quotas were assigned to the metal supply, was Alfred Arnold forced to cease production of reed boards. They were then obtained in semi-finished form from Gebr. Dix A.G. of Gera, who were mass producers supplying many German harmonica, accordion and harmonium makers. In Carlsfeld these externally produced reed boards were tuned, voiced, fitted with leather lamellae, installed and finally tuned once more inside the instrument.
Marino Rivero speaks of this instrument as a 'Stradivarius," far outclassing his other four bandoneons in its sound quality and in the precision of its response. He obtained it from the Italian bandoneon tuner Romeo Griati in Buenos Aires in 1960. All his life, this bandoneon connoisseur par excellence complained about those of his customers, mainly porteños (porteño society: inhabitants of the port of Buenos Aires), who either tried to make their own alterations to their instruments to make them as loud as possible or who brought them to him for this purpose. It was one of these porteños who gave him the future Marino Rivero bandoneon in payment, with the remark that it was unsuitable because its sound was too weak. Griati, however, immediately recognized the true value of the instrument, which had, during the first three decades of its existence, been perfectly played-in by a single owner, until it ended up in the hands of the porteño who thought it lacked power. It is a well known fact that the full tonal quality of wind instruments in general only develops gradually over a long period of time, and this is indeed true of the bandoneon. It is easy for a careless player with inadequate technique to ruin an instrument or to cause permanent damage to it. A new bandoneon of any make must therefore be correctly played-in from the start so as to allow its full tonal quality to develop gradually. Being an extremely complex combination of hundreds of individual components, the bandoneon demands the most conscientious handling on the part of the musician. The bellows only develop their pliancy and full capacity through balanced traction, the action only becomes uniformly light and sensitive through precise fingering applied to all the instrument's buttons and, in particular, the air ducts and reed boards together only gradually develop their potential sonority as a result of differentiated air regulation. Therefore it is clear that it is not only the manufacturer of a bandoneon who is responsible for its ultimate quality, but also to a great extent its first owner and player.
Griati retuned the bandoneon and changed a note, and then sold it to Marino Rivero for 1,500 Uruguayan Pesos, about U.S. $100 at the exchange rate of that time, which represented more than one and a half months' pay for a teacher in Montevideo. Griati also made the purchaser promise not to allow another tuner to touch the instrument, which accounts for the fact that this bandoneon has only been fully retuned once in 60 years, and that no work has been carried out on it that would affect its sound in any way.
When René Marino Rivero plays in an ensemble, the other instrumentalists must tune to the bandoneon's a' = 440, which is why a modern, sharp-tuned piano, cannot accompany the instrument. As far as the disposition of the keys is concerned, the manuals are in the so-called "Rheinish Layout," which was conceived in the 1850s as a 130-note model, with cyphers marked next to each of the 65 buttons. After the bandoneon had taken its place in the musical life of Uruguay and Argentina, the South American bandoneon players developed from this prototype their own 142-note model to meet their own special requirements. By adding four buttons to the righthand keyboard, they extended its compass down from c' to a, while two extra buttons on the left-hand keyboard have made the counterbasses D, E, F and F# available with either movement of the bellows. Button 11 was also changed from f#'e' (the same as button 9) to 9'f#' (fig. 4).
Marino Rivero's bandoneon departs from this standard layout only in one note, which was modified in 1960, when Griati tuned the instrument and changed the "e" on button 10 on the left manual to play "C." This is a sensible alteration because it dispenses with an identical note that is already to be found on button 6.
As the table of the two keyboards shows, the bandoneon has two notes linked to each button, one sounding when the bellows are compressed and the other sounding when they are extended. 59 buttons out of a total of 71 have a bi-sonor3 function, i.e., they each produce two different notes derived from paired tongues of differing measurements. Such an arrangement, which originated from attempts to develop instruments that were small but still had many notes, has a considerable effect on the timbre of the instrument. The formants of each note are obviously modulated by neighboring notes, so that the combination of the surrounding notes is important in determining tone quality. This sympathetic interaction has even been worked out on the basis of trial and error by some of the early instrument makers, which influenced the planning of their models. On the other hand, since the making of free reed instruments began, have also been designed uni-sonor models, which would produce the same note with either bellows movement. These uni-sonor realizations, regardless of their variants, are more systematic per se in their layouts and resulting fingering methods than the bi-sonor ones. However, the uni-sonor system in combination with the push-and- pull bellows modifies the timbre of the instrument to the degree that those players who are fixated by the characteristic tonal quality of the bandoneon, such as Marino Rivero and most South American bandoneon players, do not consider uni-sonor models to be a real alternative. They rather regard the absence of bi-sonority as a negative feature which deprives the bandoneon of its idiomatic sound.
Marino Rivero's bi-sonor bandoneon contains rows of tongues which are riveted to long plates in such a way that their free ends are exactly situated over vibration channels of the same dimensions. The smaller the clearance at the sides and free ends, the less the air loss and the better the tone. The rectangular tongues are grouped in pairs, one on each side of the plate, so that one tongue sounds when the bellows are compressed and its partner when they are extended. Eleven, twelve or thirteen pairs are combined on a long plate, each pair being fitted over its own wooden air chamber. All tongues, with the exception of the smallest ones, which produce the highest notes, are complemented by leather lamellae covering the vibration channels. They are glued at the base onto the plate's face opposite the tongue. Depending on the direction of the air flow, they are either pressed onto the plate or lifted away from it, respectively sealing or opening the vibration channels. Therefore, whenever the air stream causes one tongue to vibrate, it also opens its corresponding leather lamella and simultaneously pushes the lamella of its partner tongue against the plate, thus preventing the other tongue from vibrating.
Each vibrating tongue transmits some of its wave energy to its plate through its socket and rivet. However, "the greater the mass of the tongue carrier, the smaller the wave energy transmitted from the tongue to its anchor, and the larger the oscillation amplitude and, as the most important consequence of this, the greater the sound pressure" (Richter 1990:231).
With the long plates of Marino Rivero's bandoneon - the longest available - there is also a further factor influencing the timbre: they are made of zinc. Almost without exception, South American bandoneon players have always insisted upon zinc plates, and the German manufacturers and exporters have had to cater to this demand. This was the case in spite of the fact that in Germany during the 1930s domestic customers usually wanted aluminium, although the use of that material made the instrument more expensive. Zinc is not only heavier than aluminium, but also has a greater material density. ln the same way, the density of the steel used for making the tongues is highly relevant. Milling, pressing, stretching or rolling increase the metal's elasticity, and thus its resonance. The tongues in Marino Rivero's bandoneon have tempered segments. Today it is no longer clear what temperatures were used for the tempering and what other processes were used at Carlsfeld to increase the density of the metal. We only know that the Alfred Arnold reed boards have the worldwide reputation for having the richest sound of any ever made. Apart from their prized tone, they have the advantage of keeping stable pitch, even when played with high bellows pressure. Still, they have an extremely precise response at all times, allowing the player to produce the most sensitive and dynamic sounds.
The rule which governs free reeds is that the length and width of the tongue define its frequency. From the point of view of the tuner, this means that the pitch can be modified by the removal of material. When he shaves material away from the head of the tongue, he decreases the vibrating mass, which slightly raises the pitch. However, if he files something away at the foot of the tongue, he weakens its resilience and lowers its pitch. Tongues that vibrate at less than 150 Hz-that is to say, lower than "d"-usually have tongue weights fitted to their heads, because it is impossible to increase the length of the tongues ad infinitum without adversely altering their physical properties. Each pair of tongues has a corresponding air duct or chamber, the dimensions of which are directly related to the measurements of the tongues: for all practical purposes, their size also has to be limited.
The ducts form a row of compartments analogous to the set of tongue pairs on the long plate. Marino Rivero's bandoneon contains two types of air ducts: in the right-hand casing, shallow pans are combined with cavities that have been carved into blocks of wood that protrude from the base board (fig. 5). In the left-hand casing, only the latter type is found, with the distinction, however, that the right- hand duct floors are curved, while those of the left-hand easing are straight. This manner of installation is characteristic of double- voiced bandoneons, which make use of the so-called "Oktavton" register-i.e., each tone is produced from two tongues tuned an octave apart. South American bandoneon players favor this combination, whereas German players in the past employed every possible number and combination of voices up to five; apparently, they were not particularly bothered by the inevitable increase in the size and weight of the instrument.
Just as the type of chamber defines the air turbulence to which the tongue is subjected, so does it also determine the tongue's position in relation to the vent and its pallet, which is opened by the depression of the button. In Marino Rivero's bandoneon, the righthand easing houses tongues which are mounted parallel to the vents and pallets for the high octave and perpendicularly standing tongues (i.e., anaglott = pointing upwards) for the basic notes, whereas in the left- hand casing, both tongues are not only perpendicularly mounted but also hanging (i.e., kataglott = pointing downwards). In the opinion of instrument makers and tuners, it is this combination of different tongue positions and duct mountings that give the bandoneon its unique sound. Acoustic technicians also point to the fact that the tongue itself is not really the main sound source: this is provided by the air current which is, as it were, "blocked" periodically by the vibrating tongue, thus producing sound through quasi-intermission. The rapid motion of the tongue compels the air in its immediate vicinity to move as well, thus causing fluctuations in pressure. It seems logical to assume that the inner shape of the chambers and the position of the tongues are significant factors influencing sound amplification, radiation and tone, but this has not yet been sufficiently analyzed. In terms of its classification, the bandoneon belongs to the group of so-called intermittent aerophones, specified as free reed instruments with hand-operated bellows (Hornbostel/Sachs no. 412.132), which have no preset chords. This last fact is one of the essential features that differentiate the bandoneon from the common accordion-type instruments. Within the concertina family, the bandoneon is distinguished by its special combination of air ducts, by its long plates and by the layout of its buttons. The last few decades have seen the development of acoustic research, using advanced technology, in the field of free reeds. However, bandoneon construction has remained, up to now, traditional down to the last detail, and many unanswered questions thus remain.