Your correspondents Mr Inglis and C. M. cannot agree with Mr Hammond's theory of the circulation of hot water in pipes; and, as you remark, "it is quite evident that this is a subject that requires to be discussed." I thought Mr Hammoud's reasoning so clear that no one would have thought of disputing it, as Mr Inglis and C. M. have done. There is a small stove here, the heating of which is a practical illustration of the correctness of Mr Hammond's remarks. The structure in question is 30 feet long and 13 feet wide. It is heated by a saddle-boiler, which would do its work well enough were it not that the flow-pipe has a continuous rise all the way along one side, back the same side, round one end, and along the other side, returning back to the boiler. The rise in the 90 feet of piping is exactly 18 inches. I find, under these circumstances, that before the water at the end furthest from the boiler is perceptibly warm, the flow as it leaves the boiler is unbearably hot. Indeed in mild weather, when hard firing is not called for, the highest pipes at the far end never heat at all, while the first length or two of pipes are far too hot. Besides this, the upper sides of the pipes a few feet from the boiler are always much hotter than the under sides.

I know of another instance in which the pipes were placed in the way I have described, and with precisely the same results. Lately, however, in this case, the arrangement of the pipes has been altered, and they are placed on the level with a gentle fall for the return pipe, with the result that the heat is now much more equally distributed in the pipes, the difference of temperature at 6 feet from the boiler from the temperature at GO feet being scarcely perceptible to the hand - two-thirds of the fuel formerly required being now sufficient to maintain the requisite temperature.

In heating different houses from one boiler they certainly should, if possible, be all on the same level. Of course this has nothing to do with the matter in dispute, only it is an arrangement which has been too seldom recognised and is apparently misunderstood. It is when houses are all on the same level that the principle laid down by Mr Hammond is most easily applied.

Mr Inglis and C. M. may depend upon the correctness of Mr Hammond's remarks regarding a double circulation in the pipes when the pipes ascend. Mr Inglis compares the circulation of hot water to the drawing of water by a syphon. The two principles have nothing in common. Water circulates in hot-water pipes by reason of the different specific gravities of hot and cold water. On the other hand, a syphon acts by means of the pressure of air, and the temperature of the water has nothing whatever to do with it. Mr Inglis reasons about a heavy column of water pulling a lighter one in its wake, but cold water will run down. How, then, is it natural for water undergoing a cooling process immediately it enters a hothouse to run up-hill in a pipe in which the hottest water is at the lowest part of it? It is, as Mr Hammond says, in this case that a double circulation takes place, or the heat would never advance. Both Mr Inglis and C. M. lay great stress on heated water having a tendency to rise, and hence they consider that what is termed a flow-pipe ought to rise. Being unfortunately bothered with a heating apparatus which they consider properly fixed, I find it the best possible illustration of Mr Hammond's views.

Water is heated very slowly indeed by conduction; and I venture to say there is not an apparatus anywhere where the flow-pipe has a continuous rise in which the colder water is not returning silently at the under side of the pipe, while the warm current at the upper side is going in the opposite direction. Water in a hot-water apparatus should not be wanted to heat by convection; and if it does, there must be something wrong - and that something is a rising flow-pipe, which is wrong in principle and wrong in practice.

C. M. asserts that the point furthest from the boiler is the hottest when the pipes rise to that point. Does he mean by this that the water gains heat instead of parting with it to heat the air? The fact is, it begins to cool the moment it leaves the boiler, and the point where it enters the hothouse ought to be its highest. If it were as C. M. asserts, we ought to take the pipes a long round-about up-hill, for the sake of the heat they would gain. Not to enter further into the subject, I consider Mr Hammond's reasoning the most correct teaching I have yet met with on heating by hot water. A. H.