matter to go deep enough for the furnace. This case affords an example of heating on the level, and must be described with some detail. The houses to be heated are a lean-to and a Paxtonian; they are so nearly on a level that a quick circulation is not to be hoped for, and the lean-to has beside it a well, which frequently overflows in the winter season. Between the highest point the water reaches and the flow-pipe in the lower house the perpendicular difference is thirty inches, and between the bottom of the boiler and the level of the water the difference is only fifteen inches.
If you want the quickest possible circulation of hot water, fit a perpendicular syphon through a saucepan lid, seal the lid down waterproof and fireproof, insert a small tube at the top of the syphon, and through that tube fill the whole concern with water, and then put the saucepan on the fire. The circulation will begin instanter, and will be perfect from the first moment that accessions of heat are communicated from the fire to the water. The laws which regulate the movements of heated fluids are fully explained and illustrated in the books that treat of physics and natural philosophy, so it is not worth while to go into that part of that subject here. It only concerns us to know that in a perpendicular column of water the movements caused by heat are most decisive and complete, and that every deviation from the perpendicular tends to arrest the motion, so that when we reach the horizontal, the rate of motion is reduced to a minimum, or may absolutely cease altogether, which in a certain sense is less than a minimum, as nothing is less than the least. In the heating of these two houses we had to calculate to a great nicety how to make the most of the perpendicular space at command—to consider, in fact, how much could be done between the lowest point at which the pipes could be placed in the houses, and the highest level of the water in the well. The lower house was the only difficulty; the other was so far above it that if we ever got a fire to burn, and water to get hot thereby, on a level sufficiently low to heat the lower house, the Paxtonian would be safe enough.
Now, it is worthy of observation that the plan adopted in this case was not the only one at our command. We might have taken round the houses a flow-pipe of very small bore and a return-pipe of very large bore, and by this means have secured a circulation, and derived our heat from the return-
