The actual movement of the rod of the thermostat was only one fourth of an inch, while the complete movement of the valve was over three inches. By the use of suitable bell-joints and rods, it was not difficult to adjust the valve to the thermostat so that they would move together automatically under the heat of the sun shining through the glass.

Section of Ventilator, Chimney, Cap and Valve.

Next in importance is the peculiar heating system invented by Mr. Barnard. It is not a new idea to put an oil stove in a small plant house. The novel feature was the placing of a hot-water boiler on top of a common two-wick oil stove, and using the boiler to heat hot-water pipes. Even this idea would not be of value were not another idea supplemented to it. This is to utilize the heat of radiation from the stove, and at the same time to get rid of products of combustion. To do this the boiler was enclosed in a sheet-metal casing, and from the top of the casing a small chimney extended through the roof. By this arrangement nearly all the heat was extracted from the stove without allowing any of the smoke or gas to enter the building. The boiler was made of sheet zinc, and was 27 inches long and six inches in diameter. In the boiler were set four tubes made of one-inch gas pipe, thus making a tubular boiler. At one side, close to the bottom, was fitted a one-inch pipe, and another was fitted to the top near the center. Figure 4 shows the boiler and connections.

On each pipe were placed union couplings, and to connect the pipes with the two-inch hot-water circulating pipes reducing joints were used, as shown in Figure 4. In the figure, dotted lines show one of the flues in the boiler and the positions of the others are indicated at the bottom. A water-cock on the boiler is also shown. The upper pipe is the "flow," and the lower pipe passing down the side of the boiler to the bottom is the "return." The circulating pipes extend along the front of the house and across the end to an expansion box, open on top. This gave 38 feet of two-inch pipe, but experience showed that the boiler was able to heat 80 feet without difficulty. The pipe used was sufficient for the mild weather of last winter, but more pipe would be far better. The boiler was placed on top of a sheet-metal box 14 inches square, and having a door in front.

An opening in the top of the box, a little larger, was covered with an iron grate, and on this the boiler was supported. When in place and connected with the pipes, a round sheet-metal casing was placed over the boiler, resting on the top of the box. This casing had a tight-fitting cover at the top, and in this cover was a two-inch opening, and from this a tin pipe extended under the sash to the back of the house and then out through the post, a cap on top serving to keep out the rain.

Figure 5 shows the position of the boiler within the casing.

It will be observed that all the heat of the gas or oil flames must pass up through the flues of the boiler or along the outside. At the same time all the products of combustion are completely removed and escaped into the open air through the chimney. The boiler and casing were placed directly in the greenhouse under the sash at the south-east corner. The point aimed at was to keep within the building all the heat of the flame. Not only did the water extract a large part of the heat and carry it to all parts of the house, but the box containing the stove, the casing and chimney gave out heat of radiation directly into the house. This explains why a single gas-burner (equal to a common small lamp) kept the house warm. All or nearly all the heat was saved. With both wicks burning in the oil stove the top of the chimney was barely warm, and the hand could be put over the top of the chimney without inconvenience.

The point is just here - complete utilization of nearly all the heat (perhaps 95 per cent.) directly in the building, a perfectly air-tight building, and a continuous and unvarying source of heat. It took from two to three hours to heat the water pipes, but once heated, the oil stove kept the building warm through the coldest nights. No coal or wood fire is ever alike for more than a few moments. A gas flame or oil flame does not vary in intensity after the first ten minutes. Once filled with oil, the stove burned without attention day and night till the oil was exhausted. With a large tank such a stove would heat a greenhouse without attention for several days. With gas there was, of course, no attention needed after the flame was lighted.

Section of Boiler.

To make the heating apparatus automatic, a valve was placed on the gas-pipe, and this valve was controlled through rods and bell-cranks by a thermostat hung on the wall of the greenhouse. The morning sun shining on the glass warmed the building, and under the influence of the heat the thermostat expanded and shut off the gas, but not completely, because that would involve re-lighting the gas in case clouds obscured the sun. The flame was merely turned down low. If clouds appeared, and also when the sun began to go down, the temperature fell, and the thermostat contracted, turning on the gas again. The two thermostats worked together, one a little in advance of the other, and sometimes only one operating, as they were set for different temperatures.

Section of Casing, Boiler and Chimney.

From the experience gained, it is clear that greenhouses can be made by these simple means perfectly automatic. Gas is the best fuel, because it is most easily controlled and practically unlimited in supply. An oil stove requires a large tank to run 48 hours without re-filling, and to control the flame a special form of extinguisher must be used. Setting aside the automatic side of these improvements, the house itself is worthy of imitation, as it is portable, and because it can be operated at great economy of fuel. The peculiar form of ventilation is also a good idea, even if the valve has to be operated by hand, because it prevents back drafts of cold air and saves the heat of the sun and fire.

Mr. Barnard was assisted in his experiments by George A. Weber, of Stamford, and their inventions and improvements are offered by The American Garden to the free use of all who are interested in the use of glass in horticulture. The idea of heating greenhouses by gas or oil can also be applied to hot-beds, and it is the intention of The American Garden to soon lay plans for a new system of hot-beds before its readers.