By Jonathon Kelly
The conservation of the organ makes up a large part of the overall work of this project (see the previous posts here). Following the restoration work on the barrel, the next major problem was the leaking bellows.
This post explains how the bellows work and the problems found with them. A subsequent post will outline the treatment and associated considerations.
In order to produce music the organ needs a consistent air supply. The job of the bellows is to supply a constant air pressure to the windchest (the wooden box shown above the bellows in fig. 1), which in turn lets air into the pipes when the keys are pressed.
How the bellows work. The bellows on the organ have a large centreboard which remains stationary and is fixed into the wooden framework (fig. 2). The centreboard separates two sections; the pumps (of which there are two) are underneath the centreboard and the reservoir, which is above the centreboard.
The pumps and the reservoir connect to the centreboard with a flexible air-tight membrane. This is traditionally made from alum-tawed animal skin, a very flexible and durable material (more about this later). These alum-tawed skins are specially shaped to open and close with minimal rubbing and creasing. The pumps and reservoir hinge against the centreboard at opposite ends to allow them to work within a restricted space and to promote efficient internal airflow.
The pumps, (shown fig. 3) when opening, pull air from the outside through one-way inlet valves. As they close, the trapped air inside is pushed through another one way valve inside through the centreboard and into the reservoir above which fills with air. The two pumps work independently of each other but are worked together alternately by the organ movement.
The reservoir fills with air from the pumps and expands upwards until end of the extended arm on the spill valve (shown in fig. 4) touches the top of the wooden frame thereby tilting the valve and letting some air out.
As soon as air is let out the reservoir slightly deflates, drops down again and the spill valve closes. The reservoir then expands again and the cycle repeats itself. Under normal operation the valve opens approximately twice a second. The spill valve therefore regulates or governs the air pressure to maintain a constant feed to the windchest.
To provide the air pressure a large spring (shown Fig. 5 opened out) presses down on the top of the reservoir. In other organs the air pressure is sometimes provided by a weight instead of a spring.
The main air outlet from the reservoir is via an open airway inside the reservoir. The air exits to the windchest through a rectangular slotted hole at one end of the centreboard (shown fig. 6, left-hand side) and then through a wooden air duct inside which the bellows is fixed. The joint is sealed with an animal skin gasket.
A crank pumping mechanism driven by the organ movement operates the pumps (see fig. 7). The organ movement turns the large flywheel (as the organ plays). The connecting rod attached near the centre of the flywheel turns the rotary motion of the flywheel to reciprocal motion at the bottom via the rocker linkage which lifts and lowers the front and back pump alternately.
The problems with the Pyke bellows.
After years of use holes and tears had formed at fold points and edges which would be difficult to patch or seal (figs. 8 & 9). The growing number and size of these holes was leading to significant leakage which would only get worse if the organ was to continue to be used.
The decision was made to replace the outer alum-tawed skins rather than to attempt repairs by patching.
This process will be covered in the next post.
Inside the bellows.
Removing the outer skins revealed the internal surfaces and construction. Figure 10 shows the inside of the pumps and the air inlet valves. The valves consist of a piece of animal skin glued at each end over a rectangular hole with a flat surround to allow a seal to form. Simple and effective.
Figure 10 also shows the rectangular openings on the centre board with similar one way valves from the pumps into to the reservoir. The inside surface of the pumps and reservoir had been lined with paper.
Fig. 11 shows the inside of the reservoir with the long rectangular valves from the pumps. On the right hand side the underside of the spill valve. The open outlet leading to the rectangular slotted hole feeding the windchest can be seen between the hinges.
The animal skin.
To identify the skin used for the bellows we sought the advice of David Dorning, the West Dean books programme tutor. He identified the skin as alum taw, a traditional and commonly-used material for binding books and for bellows due to its flexibility and durability. (The tawing process is similar to the tanning process in that it "cures" the skin, but tawed skin can still putrefy if wet, while tanned won't.) The animal is likely to be either sheep or goat. The skin used inside the bellows for valves and for the gasket are more likely to be tanned.
Inside both pumps the same date has been signed in ink, March 1763 (fig. 12). One had faded or been rubbed off previously, but the other was clear to see.
The reservoir contained a watermark and makers mark on the lining paper (Fig 13).
We identified this as a watermark of a Dutch paper maker from the 18th century. The watermark is clear enough to identify a rampant lion on a platform holding a staff or similar all surrounded by a circle with an inscription (only partially readable) and surmounted with a crown. The identification is via compiled catalogues of watermarks and seems to match a watermark entitled The Lion of the Seven Provinces. From the catalogues the inscription is likely to be in Latin and probably reads Propatria Eiusque Libertate.
A makers mark of GR surmounted with a crown is also on the paper (Fig 14). These marks approximately identify the paper as being of the same period of the date on the bellows. If nothing else this provides an interesting additional piece of evidence and forms part of the Pyke's story.