i879-] HEATING BY HOT WATER. 249 



the rapid circulation of the water, but because, all things considered, it is in 

 most cases the best arrangement for other reasons." 



Now local circumstances may necessitate the sinking of the boiler below the 

 level of both flow and return pipes, but there is no reason why the flows on 

 entering the building to be heated should not be carried to their highest point 

 at once, and then commence a descending course. And I maintain that this 

 way of fixing the pipes not only "works fairly well," but that it is the right 

 way of fixing them ; and that it can be carried out more advantageously in 

 an extensive system than "where there is only one or at most two houses to 

 be heated." 



Next, Mr Makenzie says it is an error to suppose that one particle of water 

 cannot transmit heat to an adjoining particle ; but he supplies no proof of its 

 being so. In a question of the kind under discussion, assertions are useless, 

 unless backed by some means of proving their correctness, and Mr Makenzie 

 supplies no way by which it can be proved that one particle of water is able to 

 transmit heat to an adjoining particle. All experiments with the view of 

 testing this go to prove that 2) r adicatty the particles of which water is com- 

 posed are unable to transmit heat to each other by conduction. 



To prove this in a rough way, take sheets of equal dimensions of the follow- 

 ing minerals, that is, "iron, stone, lead, and ice," and hold them one at a 

 time on the palm of the hand in front of a brisk fire : the heat emitted by the 

 fire will be transmitted to the hand by the above minerals in accordance with 

 their respective conducting power, the surface of the ice exposed to the action 

 of the fire will receive as much heat from the latter as the "iron, stone, or 

 lead ; " but because of the inability of the particles of water to transmit heat 

 by conduction, the palm of the hand is not warmed, but, on the contrary, is 

 exposed to a temperature at the freezing-point so long as the thinnest film of 

 ice intervenes between it and the source of heat. Yes, " water is water," and 

 for any practical purpose has "no power to transmit any heat by conduc- 

 tion." Then about the return current in the flow-pipes, Mr Makenzie admits 

 that it takes place in the first instance, but says, "The back motion in the 

 flow-pipe must immediately cease as soon as the average temperature in the 

 flow-pipe becomes higher than the average temperature in the return." Now, 

 if this were correct, "the back motion" would not continue more than a few 

 minutes after the fire was set agoing below the boiler. As the first volume of 

 heated water that entered the flow would raise the average temperature of the 

 latter above the average temperature of the return, so that back motion would 

 not be worth talking about if it only continued the length of time indicated 

 by Mr Makenzie. It continues, however, and is as constant as the forward 

 motion in all apparatuses that have the flow-pipes laid on a continuous ascent 

 throughout their length. The reason for the continuance of the return 

 current is plain. The point of the apparatus on which the fire acts contains a 

 volume of water of less specific gravity than the lower strata of water in the 

 flow-pipes, and whether we treat the question as one of "pure hydraulics" 

 or as one of "hydrodynamics," it won't alter the fact that water of less 

 specific gravity cannot force that of a greater uphill, and therefore the colder 

 and relatively heavier volumes of water in the under side of the flows will roll 

 back in hotter and relatively lighter volumes in the boiler, "like so many 

 boulders tumbling down the Cumberland Screes," notwithstanding what Mr 

 Makenzie may think to the contrary. 



Again, Mr Makenzie says, " The difference of temperature between the upper 

 part of the flow-pipe and the under part has nothing to do with it" (that is, with 



