In his third patent of 1881 Sax speculates on possible changes in the mechanism of the saxophone. He outlines options that might seem odd to us, but which certainly are not illogical. Perhaps Sax is inspired by the oboe, but in any event he gives a couple of interesting solutions for the fingering of semitones.

Since at that time Sax was no longer the leading maker of saxophones, little came of it. The mechanism had meanwhile developed in another direction toward the form in which we know it now. And all things considered, conceptwise very little ingenuity has been put into it since, say, the beginning of the 20th century. Yet Sax had different plans.

The drawings alongside stem from the third patent and give two different possible layouts of the mechanism. Both bell and palm keys have been omitted, of course. Please note that in the drawing on the right the pink and purple coloured keys nrs 10 and 12 do not have the same function as the present side keys for Bb and alternative C; these keys have the same function as their counterparts in the drawing on the left, but in another layout.

Yet there is an important difference between both mechanisms: in the drawing on the right the third finger left hand operates the key for G–G# and not the key for G#–A. Other than we are now accustomed to, this key is here an open key! In the drawing on the right the habitual function for G–G# for the little finger of the left hand is missing and G# is taken more as an Ab by closing key nr 10 with the first or second finger of the right hand! This very same function closes keys nrs 12 and 14 as well and the result is, that with the right hand we can obtain a semitone flattening for any fingering of the left hand.
It is interesting too that Sax introduces a 'speaker key' which is coupled to key nr. 9 to ensure an open sonority when this Ab is taken. Key nr. 6 or 7 and 8 namely close at the same time and would otherwise muffle the sound of this Ab.

Another important distinction between both proposals for the mechanism and the conventional type, is the way in which C is obtained. The 'recorder–fingering' (0-2) is not applicable. In stead we have to use the above mentioned flattening by a semitone with the right hand and then, this time, applied to the fingering 'everything open' (C#). The fingering for C thus becomes 0-0-0=0-5-0, or 0-0-0=4-0-0. In fact this is a very consistent solution. Maybe a bit too consistent even, as we are not inclined to think of a diatonic C as a lowering of a C#. In daily practice, a saxophonist will not easily be prepared to switch to this type of fingering, although in the oboe it is quite common.

Further note that in the drawing to the left Sax did not insert a P-function for the Bb. He might have done so. There is a patent for this mechanism which is held by Evette-Schaeffer, but that dates only from 1887. Until that time obviously no one has grasped the opportunity.

The Londeix system for defining fingerings is used.

in practice: serial number 40842

It seems that Sax built only one instrument which incorporates the ideas of the third patent: alto saxophone 40842. Wholly in accordance with the patent, the range of the instrument up high is extended up to G3, but the layout of the key spatulas is somewhat different from the one we're used to. On the other hand, the mentioned extension in the low range toward Ab0 is not there. This can be because the proposed layout of the mechanism would in practice prove to be unplayable, or, as is possible as well, because the instrument's lowest notes proved to be unstable. There is an extra key for a 'short' middle range D2, such as was added frequently by german saxophone makers later on. And most notably, the extra register keys that are described at some length in the patent were not added, but for the rest, the instrument is a brilliant example of intelligent self-criticism, where Sax circumnavigates the maybe harder to accept proposals of the patent with a mix of both mechanisms and some minor shifts in layout, in such a way that

  • the mechanism provides a kind of a Tf-function (alternative F#) without adding an extra key.
  • there is both the flattening of A to Ab with the right hand (though without the speaker key that was foreseen in the patent), and the function for the little finger left hand for G–G# (directly operated and not yet linked to the right hand mechanism).
  • there is almost a P-function for the Bb, but the design of this part of the mechanism is not entirely clear, maybe unfinished in some detail. Otherwise, we find the usual Ta-function (side Bb) and the possibility to obtain Bb with the right hand as a semitone flattening of a B.
  • there is a possibilty to obtain a C in the usual way (0-2) and there is the alternative to take the note as a flattening of a C# with the second finger right hand.
  • Only the keys numbered 9a and 11a are closed keys.

    It's a pity, for the further course of history, that this instrument didn't play a role of any significance whatsoever.