1835.] Description of a self performingbarr el organ. 35i 



respect to small machine organs, because, it is so much 

 of the moving power and of the air generated expended 

 to no useful purpose. And as the same piece of mecha- 

 nism usually performs both the office of blowing the 

 bellows, and turning the barrel, it occurs in many cases 

 that the music is of necessity played faster than it 

 ought to be, especially in songs, because the bellows 

 do not generate air fast enough : and the slower the 

 music the greater the quantity of air consumed. 

 Now one piece of mechanism v»dth only one first moving 

 power cannot effect two opposite motions at the same 

 time, namely, a slow movement for the music, and a 

 rapid movement for the bellows. — This involved what I 

 thought to be a defect in design which I obviated by 

 employing a separate piece of clock-work for working 

 the bellows ; and in order that the air generated might 

 not escape and be wasted, I made the reservoir without 

 a safety valve, and provided that the apparatus should 

 regulate itself, never generating more air than the music 

 required, yet generating sufficient for a full supply, how- 

 ever slow the music might be, or however crowded with 

 chords, and counterpoint. — Had not some expedient of 

 this kind been resorted to, the weight F in figure 1, 

 would have descended quicker than the weight E, and 

 probably before the weight E had descended two thirds, 

 the weight F would have sunk to the bottom, and have 

 left the barrel movement and music without air : but by 

 the contrivance adopted, the descent of the two weights 

 are now as equable and perfect, as the nature of the in- 

 strument admits of. 



14. — Figure 9 shows the arrangement ; but it is 

 drawn to no scale, and exhibits only ihQ principle of the 

 jnanner in which the bellows movement is regulated. 

 Let L represent the reservoir ; a, a, small rollers or pul- 

 lies over which the line b, b, passes, one end being fast- 



