From the drawings it may be seen that the pit is prolonged to the side wall of the building, with steps leading up to the ground-level. The top and sides of the opening are made with galvanised sheeting, forming a kind of short tunnel in which the coolie may stand upright. The outer face of the brick furnace forms the inner end of the tunnel, and accommodates the door of the furnace. The bottom of the pit is filled up with clay and stones almost to the level of the bottom of the door. This ensures a very shallow hearth, and guards against an unduly large fire. Obviously it is not desirable or necessary to make the hearth of fire-bars, as was done in one instance, with the provision of a door below for removing the ash. This would lead only to a strong draught being created, with a high rate of fuel consumption. [Illustration: BRICK AND CEMENT SUPERSTRUCTURE OF FURNACE INSIDE THE BUILDING, BUT FED FROM OUTSIDE. On the top of the superstructure rests a sheet of perforated boiler-plate. The actual fire-pit is below ground-level, and to the left may be seen parts of the sides and top of the downward approach, from the outside, to the door of the fire-pit.] The openings can be screened by a narrow sloping lean-to, which serves to keep out rain, and provides shelter for the stock of fuel and the coolie. The iron furnace-door should be well made, with an easily worked latch; but it is not necessary that it should be perfectly fitting. Any slight aperture will serve to provide the necessary air-inlet, but in any case it should not be more than slight. CONSUMPTION OF FUEL.--Regarding this furnace, it may be said in conclusion that it is more satisfactory in general working than any other furnaces yet encountered. Obtaining information from over sixty estates, on the question of fuel consumption compared with output of rubber, it was found that, as far as ordinary smoke-houses were concerned, the Pataling type of furnace showed the lowest unit consumption of fuel. [Illustration: GENERAL VIEW OF SHELTERS COVERING APPROACHES TO FURNACES.] This was at the rate of slightly less than 1 lb of fuel per 1 lb of thoroughly cured sheet rubber. The figure on some estates mounted as high as 4-1/2 lbs. of fuel per lb. of rubber. Naturally this factor may have been affected by failure to utilise the drying space to its fullest capacity, but in the main the high rate of consumption could be attributed solely to the deficiencies of the furnaces.