ON THE CONSTRUCTION AND HEATING OF HOT -HOUSES. 
253 
proposed for the purpose, especially in those for employing oil and other fluids, at a 
temperature considerably above the boiling point of water. As the heated air and 
vapours must necessarily enter the chimney at a higher temperature than that of 
the fluid in the boiler, the loss of heat will be greater in proportion as that tempera- 
ture is increased. 
There is another circumstance connected with the plans alluded to that does not 
appear to be taken into consideration at all — the great capacity of water for heat, 
compared with most other fluids, oil for instance, or mercury. Water, it is well 
known, in being heated any number of degrees, absorbs twice as much caloric as an 
equal bulk of oil or mercury, and consequently a given bulk of . . • . F ig. 4. 
water at 212 degrees, in cooling down to 60 degrees, that is, in 
losing 152 degrees of sensible heat, would warm the surround- 
ing atmosphere as much as an equal bulk of oil or mercury 
would do in cooling from 364 degrees to 60, or in losing 304 
degrees of sensible heat, as indicated by the thermometer. 
Such being the case, if oil or mercury, or almost any substance 
that can be named, could be compared in point of economy 
safety and cleanliness as a material, with water at or under the 
boiling point, still it would be inferior to water in point of 
economy as regards the consumption of fuel. 
The apparatus represented by the figures 1 and 2 is constructed of 
copper, and costs about 91. ; a house thirty feet in length would not cost 13/. In 
the plans 3 and 4 the outer case or boiler is supposed to be cast iron, as 
well as the case of the furnace in 4, and all the other parts in copper. The 
whole might be made of cast iron- ; but there would be a risk of its cracking in those 
parts exposed to the fire, an accident to be especially guarded against in heating 
horticultural buildings. Plate-iron or tin-plate might be employed, either wholly 
or in part ; but cast iron and copper would perhaps be preferable materials. With 
respect to the pipes, when made of sheet-copper or zinc, or tin-plate, they are better 
calculated to disperse the heat than those of cast iron commonly used, which are 
perhaps, except in point of strength, the worst that can be employed ; they expose 
a smaller surface, in proportion to the water they contain, than pipes of any other 
shape ; and from the thickness of metal, never less than three-eighths of an inch, 
and frequently more, they oppose a direct obstacle to the ready transmission of the 
heat from the water to the surrounding atmosphere. 
Thin sheet-copper is certainly far superior, and perhaps, in the end, the most 
economical. Zinc, though otherwise an excellent material, is not well suited for 
elliptical pipes ; it becomes so soft at the temperature of boiling water that they 
soon, the lower ones especially, change their shape and become nearly round. If 
cast iron be preferred, there is no reason why the cylindrical form should be adhered 
to ; and probably if elliptical pipes of different sizes, cast as thin as possible, were 
to be supplied from the foundries where those now in use are cast, they would soon 
supersede the use of the latter. 
No reservoir is attached to the small apparatus I have described ; but on a large 
