324 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
and the hot water is forced upwards, thus the circulation is begun and 
in the same manner continued. It is clear that the cause of the upward 
flow of the hot water in the flow-pipe is the greater density and weight 
of the higher and colder column of water in the return -pipe. 
It can be proved that with an apparatus having a height of 5 ft. from 
the lowest to the highest point, and with an average difference of 
10 degs. between the flow- and return-pipe, the water in the return is 
continually falling with a theoretical velocity of 68 "4 ft. per minute. 
With an average height of 10 ft. the fall per minute is 96*6 ft., and in an 
apparatus having a height of 20 ft. the theoretical fall is 136-2 ft. per 
minute ; in short, the motive power in a hot-water apparatus is entirely 
in the return-pipe, the amount depending on the height and on the 
difference of temperature between the flow and return. 
In quoting these figures no account is taken of the friction, which 
may be very small, or may be sufficient to wholly stop the circulation. 
In the heating of hothouses it is of the utmost importance to re- 
member that the motive power is in proportion to the difference in height 
between the lowest and the highest points of the apparatus, which prac- 
tically means the depth of the stokehole and rise of the pipes. Attempts 
are sometimes made to avoid sinking a stokehole, but such attempts are 
bound to be failures, and are only attempted by people without any 
knowledge of the underlying principles which govern the circulation of 
hot water in pipes. 
It is well known to all experienced heating engineers that a boiler quite 
powerful enough to heat 1,000 ft. of pipe where there is a height of 25 ft. 
or 30 ft.will not efiiciently work more than 750 ft. when the height is only 
5 ft. or 6 ft. 
Along with this must be considered the frictional resistance, which is 
the work to be accomplished. 
For hothouse work there is a general agreement that a 4 in. pipe is 
the most suitable in regard to the quantity of water and the friction on 
the walls of the pipe ; 3 in. and 2 in. pipes may, and often are, used, but 
probably 80 per cent, of the hothouses erected are heated with 4 in. In 
very large apparatus larger pipes are often used for mains, but the 
radiating pipes are almost invariably 4 in. 
The relation betw^een the size of the structure to be heated and the 
amount of heating surface is of the greatest importance ; and although 
there are no scientific rules for this, practice has been much on the 
following lines, which, I think, except in the most exposed situations, are 
safe lines. Of course provision must be made against the coldest weather, 
which may be taken at 32 degs. of frost. 
For conservatories where a temperature of not more than 45 or 
50 degs. is wanted there should be 1 ft. of 4 in. pipe, or its equivalent, for 
€very 35 cubic ft. of space. 
For plant-houses, w4iere a higher temperature may be required, the 
proportion should be 1 ft. of pipe to every 25 or 30 cubic ft. of space. 
For stoves and orchid-houses, and also for early vineries, the propor- 
tion of heating surface should be still higher. An orchid-house 12 ft. wide 
requires four rows of 4 in. pipes along each side, which gives 1 ft. of heating 
surface to every 12 or 13 cubic ft. to be heated. 
