This section is from the book "The Gardener V1", by William Thomson. Also available from Amazon: The New Organic Grower: A Master's Manual of Tools and Techniques for the Home and Market Gardener.
The interesting discussion now taking place in "The Gardener' on heating by hot water will be the means of showing to a great extent whether deep stokeholes are a necessity or not; and I am in a position to coincide with Mr Hammond that deep stokeholes in many cases are not required. "We have in use tubular and saddle boilers, with a hermetically sealed box placed nearly above them, and into this box the hot water flows from the boiler, and then falls perpendicularly 7 feet to the pipes, which are then level throughout, the return pipe only falling 1 foot gradually at 9 feet from the boiler. It will thus be seen the only rise in the pipes is immediately above the boiler, and the returns fall at 9 feet from it - one boiler alone heating five houses, comprising an early vinery and stove; and during the past winter the heating arrangements were quite satisfactory. It is well to state that each range of houses is built on a level, and it is only in such cases this system could be satisfactorily carried out.
But this system in our case proves conclusively to my mind that a continuous rise and fall is not required to cause circulation of hot water in pipes, and only where houses are built on different levels is it necessary to have the pipes highest at the extreme point, and in such cases the return pipe is high also. The theory that the higher the extreme point of the flow-pipe the more rapid the circulation of water in them, to my thinking is an error; for however high the elevation, the water would remain stationary without some other agency to bring about circulation, and the main agent is heat, and the more heat that is applied the quicker the flow. To further convey my meaning, let us commence with a fire just lit, and the boiler and pipes filled with cold water - that is, of equal density. The water in the boiler exposed to the heat will gradually get more rarified, and is displaced by the denser body which rushes in and takes its place, which is also made lighter and is again displaced, causing suction along the whole length of piping, drawing the water onwards in its wake, - and shows that however high the piping at the extreme end, the real centre of circulation is in the boiler, and the quicker the water is heated the quicker its displacement and the stronger the suction, and, as a coo-sequence, the more rapid is the circulation of water in the pipes.
I would here point out also that the higher the ascending pipes, the more suction is required by the descent to draw the water from the ascending pipes, and in this way the high elevation is equalised to a proportionate degree. As I have based my argument on suction, and it brought about by displacement of the lighter by a heavier body in the boiler, I may be allowed to show that it is suction that causes the circulation of hot water in pipes, and this can be conveniently exemplified by drawing a portion of the water off, leaving a vacuum at the highest point: the remaining body of water will then find a common level, and circulation in the pipes will cease. This shows that if the fire forced the heated water up the flow-pipe, it would continue to do so until the returning pipes were empty. But that this is not the case will be evident to all who will try this simple experiment; for the water will keep a common level, and, fire away as we will, this state will not be altered, but the water will get heated slowly by conduction in the flow and return, and at an equal rate: but again fill up the pipes, and there is displacement of the rarified body by one of greater density at a slightly lower elevation, renewed by suction in the pipes, and causing circulation.
Further, to show as a siphon that a high elevation is no promoter of quicker circulation, take a circular vessel, A, 4 feet deep, 2 feet in diameter, and standing 4 feet higher than B, of equal size; fill A full of water, and then bring your siphon 12 feet high or more from A to B. Let C and D be vessels of an equal size to A and B. Also, stand C 4 feet higher than D. C being full of water, bring your siphon only 2 feet high, the siphon reaching the bottom of each vessel. Let the water run in both siphons at the same time, and 2 feet will be drawn from A and C to B and D in the same time when the water in the vessels will have found a common level, and the siphon will cease to act, showing that no circulation can take place until the water in the boiler is made lighter and is displaced by the weightier column, and that it is not necessary to have a continous rise or a high elevation to bring about this result. Albion.