This grand book should be in the hands of every one thinking of doing his own heating. My good father sent me a copy thirty-seven years ago, but I loaned it to a plumber and steam-fitter who has long since retired to an atmosphere where steam or hot water heating are little discussed, and I suppose he took my book with him. If so, that copy is decidedly out of print.

Now this illustrates the motive power that first starts the circulation of hot water. It is the difference between the weight of the water in the return pipe and that in the boiler. The water in the boiler being made lighter by the fire, the colder and heavier water forces it up and it is replaced with cold water, so it must follow that the higher, and consequently heavier, the column of water in the return pipe the faster will be your circulation. And it follows again that the faster the circulation the hotter will your pipes be, for the water returning quickly to the fire has not time to get cool. "When your return pipe near the boiler is nearly as hot as the flow where it leaves the boiler your circulation is perfect. All of which goes to prove that the lower the boiler the better the apparatus will work.

Reserve all your drop till you get near the boiler and then drop perpendicularly down. This talk about giving the pipes a rise of one foot in a hundred, or the same drop, is all bosh. If the pipes were a dead level in the house it Mould be perfect, but it is better to have a rise or fall of two inches in a hundred feet because you want when emptying the pipes to have a drain out. Providing your boiler is well down, and that is the very essence of the whole job, it makes no difference whether you have a slight rise in the flow pipes in the house or a slight fall.

As admitted at the start of this article, I was the victim of false theories and it took a few years to break away and return to principles I knew of many years ago. Now here is what we did some fourteen years ago and it was wrong. The top of the heater was some two feet six inches below the joists of the boiler shed. It would have been lower could drainage have been secured. As it was we had to lay a 4-inch socket tile 650 feet. The two 3-inch flow pipes rose straight up through the floor and eight feet higher. This was done to avoid the doorway leading from the shed into the greenhouses. Please imagine these two houses to be 20x125, running east and west and connected by wooden partitions and wooden gutter. The flow pipe in the houses was a 2-inch on each side of the houses, running a foot below the gutter and on the sides, on the posts a foot below the plate from which the bars spring; that is a good place for them, but it is by no means overhead heating, it is resisting the cold at a very important spot. As we had carried the flow pipe from the boiler so high in the shed, we had to drop again to the level of the flow pipes in the houses, which was done with two elbows and a short piece of 2-inch pipe. Although it worked, this was against correct principles. The hot water had no natural tendency to descend, it would rather flow out on a level. At the farther end of the 125 feet the 2-inch flow went into a manifold with five 1 1/4-inch openings and they were the returns. This was not anything like as wrong as the case described earlier, yet there was not hot water enough in the 2-inch to supply the five 114-inch. That and other causes, friction and rapid radiation, all combined made it a poor job. There was no need of air pipes or petcocks in this system because a 3/4-inch pipe was tapped into the highest point of pipes in the shed and led up to a large tank which supplied us with water for the houses. This tank held 150 barrels of water and from it a 1 1/4-inch pipe was connected with the return pipe close to the boiler. It may have given us a constant pressure of fifteen pounds. Please don't attach the slightest importance to the fact that this feed pipe was from a vessel holding 150 barrels of water. Pressure in a liquid depends entirely on height, not bulk. A funnel thirty feet high and three feet in diameter at the top, tapering to one inch at bottom, would give no more pressure at the one inch opening than a column of water one inch in diameter and the same height as the spreading funnel. Now we ran along several years with the arrangement above described, thought it perfection, and read that others in the east had adopted the same principle. Yet now we feel sure that many tons of coal were wasted in trying to get those return pipes hot. With the small amount of radiation we wanted them hot, not lukewarm.

So two years ago last summer we made a great change, not with the slightest feeling of doubt or uncertainty, because we were only going on the sound laws and principles of hot water circulation. e needed a new heater, wrongly called boiler, and purchased the Burnham. Let me say here that I have reason to deeply regret recommending boilers in the past without sufficient experience with them. Still, after two years' trial I must say that the Burnham boiler has proved most satisfactory. It is of heavy castings, simply put together and easily and thoroughly cleaned. In place of one 2-inch flow pipe we hung two 2-inch and at the farther end ran each flow pipe into a manifold with three 1 1/4-inch openings. The 2-inch could amply supply the three 11/4-inch and they were hot back to the shed. I should have started at the boiler. Will say now that the Burnham boiler, or the size we have, has a 5-inch opening for flow, same for return. We ran up as near as possible to the joist with a 5-inch pipe, then branched with a tee into a horizontal 4-inch pipe. I now remember there were eight 2-inch pipes to be supplied from this 4-inch. On reaching the south side of the north house we branched with a tee into a 3-inch and rose to the level of the flow pipes. Remember, the 4x3x3 tee came out of a 4-inch horizontal, not an upright, pipe. After supplying that run the flow was reduced to 3-inch and went on to supply the north run of the north house. From the shed end to the farther end the 2-inch flow pipes have a rise of two inches. On reaching the manifolds provision had to be made to let air escape, so we tapped into each manifold a small petcock. For the first ten days after we begin to fire with a new supply of water there is a considerable accumulation of air. After a week or two there seems no more in the water, and if these petcocks are neglected for weeks no harm results. This I have found repeatedly to be the case. You may not have a tank of water raised up thirty feet, and there is no need of it. A barrel or box holding twenty gallons of water, or a metal boiler, such as is used in our kitchens, will do equany as well. Let it be seven or eight feet above the bottom of your heater, and when there is a gallon of water in the barrel your system is full. In this case you do not need petcocks, but in place of them simply tap in a 1/2-inch pipe and let it run up a rafter or anywhere out of the way; that, of course, will constantly and with-, out watching vent the heating pipes of any air. See that the top of this little pipe is a foot or two above the highest point of the water in your feeding cistern. This rearrangement of pipes was a great satisfaction and unqualified success, so much so that in seven or eight other houses where the overhead and rise and fall delusion had existed it was all pulled down and common sense and a smooth, quick circulation exists.

If an amateur came to me and asked for advice about heating a house 20x100, or 20x75. and wanted it so arranged that he could leave it at 11 p. m. till 6 a. m., I should without hesitation tell him to get plenty of boiler power and use the 4-inch pipe for radiation. Use the heavy cast iron pipe used by horticultural builders, 9-foot lengths and four inches outside diameter, not what many greenhouse men have used, Mott's soil pipe, 5-foot lengths, four inches inside diameter; this is a thin, brittle metal, unsatisfactory in many ways.

I have had no experience with water under pressure, unless the pressure of our city mains constitutes that system. Our city water has a pressure of about thirty-five pounds to the square inch, and in my experience it is a very cheap heating system for a store or office. Wishing to heat a flower store in this city, which is something like 19x80 feet in area, I put a small heater in the cellar. It is simply three lengths of 3-inch pipe, each about three feet long, and run into a manifold at both ends. The coil is resting on two 4-inch brick walls about two feet from the floor and is bricked over top, sides and ends. One end of this coil is raised about three inches and from it rising to the floor is the 11/4-inch flow, which leads off, and by the help of some tees connects with three radiators on the floor of the store, and from the other end of the radiators the returns drop to the lower end of the coil. There are two natural gas burners under this very simple heater, which in the coldest weather has never been turned on more than one-third its force.

A 1-inch pipe from the city water is connected with the lowest part of the coil and the valve is never closed, so there is always a pressure of thirty-five pounds on the pipe and radiators. The highest part of the system is the top of the radiators, and in them is a petcock which should be opened every day to let out air, but often is not for weeks, and in a radiator it is not of so much consequence.

There is nothing more about it, only the radiators can be made very hot; a great success. If a strong fire should expand the water in the heater it has to find room by driving the water back into the mains. The whole thing cost less than $50, and $5 worth of gas was consumed in the coldest month. Now this system could be used with great success wherever you have a boiler that would stand the pressure. You could use it on either the uphill or downhill systems, but you could not have any open air vents, and unless you trusted to the automatic air valves you would have to daily open the petcocks at the highest point.

The architecture of many greenhouse establishments, the smaller and older places more especially, is so complicated and houses are on so many different levels, that sound advice as to how to arrange pipes is impossible unless you are on the spot. Three rules can be accepted as sound. Provide plenty of capacity in both flow and return pipes. Don't use less than 2-inch pipes, and reserve all the drop on your return pipes till just before you enter the boiler. Watch these points and you will save coal.