Langdorff serials 20303 and 20430

An excellent study by Bulleid: Langdorff serials 20430 and 20303

An interesting Langdorff box made in 1881 is Serial No. 20303. Although it has lost its tune sheet it was probably labelled Harp Harmonique Tremolo. It is a six-air movement with a 13“ (33 cm) cylinder. The main comb, at the bass end, has 88 teeth including groups of six, five and four teeth of the same pitch for the mandoline effect. The treble comb has 35 teeth and the pich overlap is such that the highest twelve teeth on main comb cover the same range as the lowest eight teeth on treble comb. The overlap results in one note having eight teeth. Fig. 1-14 shows this comb arrangement compared with the Concerto Tremolo arrangement of Serial No. 20430. Both have zithers mounted from the bedplate and covering most comb teeth; it is not possible to be certain how these were originally arranged, but I found it most effective to limit the zither tissue to the tremolo comb on Serial No. 20303. The zither holders were decorated with various engraved musical motifs.
I had to repin Serial No. 20303 so I can report that it has 5,729 cylinder pins; 2,098 for the main comb and 1,631 for the treble. By what must be pure coincidence these both average 46.6 pins per tooth. The average playing rate throughout the six tunes is just under 16 notes per second which is in the normal mandolin range. As usual on harp Harmonique movements, the stiffness of the teeth is the same in the two combs. The tunes include the ToreadorĀ“s Song (1875) and three waltzes. The waltzes display an effective trick not often used by tune arrangers of a slight reduction in tempo to emphasize the finale. It certainly made a lot of sense for Langdorff to produce both of the types shown in Fig. 1-14. They are quite distinctive, one with the richness of sublime harmonie and the other brilliant treble or piccolo effects, and both with mandolin decoration.
The Gamme number of Serial No. 20303 is 192; but Langdorff Gamme numbers, under suspicion for some time, are now rather discredited as useful data because three identical Concerto Tremolo boxes have different numbers. That will take a lot of explaining. However, I think it is well worth recording Gamme numbers in the hope that the eventual explanation may shed more light on Langdorff procedures.
(Bulleid, Technology, p. 19-21)