http://www.hevac-heritage.org/items_of_interest/heating/other_buildings/botanic_gardens_belfast/botanic_gardens_belfast.htm
splendid info about the Ravine and Palm House heating!!

THE WEBSITE OF THE BUILDING ENGINEERING SERVICES HERITAGE GROUP
Heating System discovery in the Tropical Ravine House.
The existing heating system is provided by banks of 4-inch cast iron pipework coils which are routed at high and low level around the glazed perimeter of the building. During investigative works carried out recently to discover the source of leak/s which were causing the constant loss of water from the system it was found that another circuit of pipework had been run inside a perimeter brickwork duct which had been purposely
covered with soil. This duct when an entry into it was opened up was found to start in height as high as a walkway but then gradually decreased in height until it became a crawlway before finally ended as two adjacent earthenware hollow pot air passageways . At this point the cast iron pipework left the duct and was then buried in the ground. Through the passage of time the wet soil had corroded the surface of the cast iron and ultimately the pipe wall which eventually failed and created the leaks.
The discovery of this duct brought to light an earlier method of heating the Tropical Ravine House. The duct had originally been used as a horizontal path for the flue gases to escape from an open hearth fire which was kept constantly fired in a basement room. The flue gases supposedly warmed the walls of the brickwork duct and the soil piled around the duct, which provided a continuous constant heat flow into the Ravine House. The duct originally had been used as a horizontal path for the flue gases to escape from an open hearth fire which was constantly kept fired in a basement room.
The gradual reduction in size of the flueway kept the flue gases moving at a slow enough velocity, so that they could distribute and transfer their heat into the surrounding brickwork structure. At the end of the duct the flue gases finally passed into the two earthenware pot shafts before connecting into a vertical brick chimney at the other side of the building from the fire room.
It must have soon been found that this method of heating was proving unsuccessful as the long horizontal duct run for the flue gases could not provide sufficient draught in the system to draw the gases into the chimney. We can only imagine what the room conditions were like for the unlucky stoker trying to keep the fire well alight.
That this original system did not prove successful is evident by its replacement with the more reliable method of using 4" cast iron hot water coils as the heating system.