I must preface this very important subject by a confession. Some fifteen years ago I made the acquaintance of a steam-fitter, a clever workman, who had been quite successful in heating dwelling houses and thought he had ideas on greenhouse heating, but on that he was away off. A weakness of the writer is to be anxious to learn from any one who he thinks knows more on any subject than he does. So I learned from this man several fallacies on greenhouse heating which I have many times regretted putting into print. It does me good to confess this and I must ask the reader of the article on hot water in the first edition of this book to believe that where my present views conflict with what I wrote six or seven years ago, it is the result of later experience and practice. One of the fallacies was the use of small pipe, that is, less than 2-inch. Looking at this question superficially, the small pipe, 1-inch or 1 1/4-inch has the advantage over larger because you get more radiating surface in proportion to volume of water, but in practice this does not hold good. Friction is so great that the circulation is retarded, radiation is rapid and the small amount of water in the small pipe soon cools. While 11/4-inch pipe may work very well in a house not over fifty feet, for any longer distance use nothing less than 2-inch.

The next and most costly mistake for me was running the pipe straight up from the boiler perhaps to within a foot of the greenhouse ridge, using a 2 1/2-inch or 3-ineh pipe, and at the farther end then dropping into a number of returns containing eight or ten times the volume of water of the flow pipe. Here is a dismal specimen of one of the houses, and I have reason to believe the same mistake has been made in hundreds of cases.

He had a 4-inch pipe rising from the boiler to the height of the greenhouse ridge, then branched off with a 3-inch and suspended the 3-inch about a foot from the ridge to the farther end. The house was used for soft-wooded pot plants, and was 20x100. At the farther end the 3-inch branched into two 2-inch pipes and dropped to the level of the 4-inch cast iron pipe under the benches, with which the house was formerly heated, four under each bench. By a complicated lot of fittings one of the 2-inch pipes was actually expected to heat the four 4-inch. Could anything be more absurd? Roughly estimated, the four 4-inch contained sixteen times as much water as the one 2-inch. Now this was an extremely absurd case, yet I have seen many others, where the same mistake exists differing only in degree. What were the principal features in this failure? The pipe near the roof at the boiler end of the nouse was very hot for perhaps twenty feet; then it gradually cooled and for more than half the length of the house there was not heat enough for any to be diffused among the plants on the benches. When the 2-inch pipe emptied into the four 4-inch returns, 16 slightly warmed them for a few feet, and then for seventy-five feet there was no-heat under the benches.

I must digress a moment here to say that although growers generally want no steam or hot water pipes beneath their benches of roses or carnations, they are of undoubted benefit to our soft-wooded plants like geraniums and begonias. Every one who grows bedding plants must have noticed this, but under the arrangement described we had little but cold, damp benches. There is a time of year, fall and spring, when this pipe near the ridge is useless. Days or nights when it may be 50 degrees outside you need a little fire heat to keep out dampness and keep up a healthy circulation of the atmosphere. While firing gently you will also want a little ventilation, and the heat from overhead pipe will entirely be wasted and your house left chill and damp.

There was an absurd controversy in the florists' papers some fifteen years ago on this subject of overhead heating. Some said that as the natural source of the heat was from above, meaning the sun, that our pipes should be near the roof, a great fallacy. Another piece of information which appeared some years ago in answer to a query was, " If you use hot water you need not excavate for heater, if you intend to use steam then you must keep the boiler well down below the level of the greenhouse." Nowadays we know that, broadly speaking, the very reverse of this is the truth, for it is the return, or cool, water pressing on the warmer and lighter water in the heater that is the power or cause of circulation, therefore the greater the height of the descending column of water the faster the circulation. In truth there has been nothing new discovered of late about the circulation. There may have been in its application.

Some forty-five years ago, perhaps before, there was published in London by Hood a volume on hot water. There has never been a better work on the same subject since. We may have found out better and cheaper modes of applying it than prevailed in his day, but all the laws of circulation which he demonstrates so finely are just the same today and always will be, for they are natural laws, and can never be altered. Hood says that the circulation of hot water was well known by the Romans, and used for heating their baths, so this wonderfully useful method of warming our houses did not originate in London, New York or Kalamazoo.

The cause of circulation is finely illustrated by Hood by having two columns of water, say two lengths of 4-inch pipe, each three feet high and connected at the bottom with a piece of pipe of the same size. Half-way in this connecting pipe there is a valve. Fill one pipe with water at the temperature of 45 degrees, the other with hot water at a temperature of 180 degrees. Then open the valve connecting the hot and cold columns. The hot water will rise or overflow. Now this is all there is to the circulation of water in your system. The water in the cold column or your return pipe is denser and of greater specific gravity, in simpler words, heavier and forces the lighter and warmer water out through the flow.