IRON AND WOOD ROOFS FOR STOVES, &c. 
35 
the lights, as to admit the external air in a degree sufficient to counteract entirely 
the power of two strong fires, when the flues have been heated to the greatest 
excess, before the temperature of the house could be raised to 30°, the ther- 
mometer then standing (out of doors) at 18° of frost : this was in February, 
1830. Now this occurrence took place in a house of no very large dimensions 
compared with the wood-roofed vinery I am about to describe. The dimensions of 
this building were forty feet long by sixteen wide and nine feet high, with a pit in 
the middle for the culture of pines, &c., which very much reduced the cubical 
number of feet of air to be rarified, compared with the wood-roofed house, which 
was fifty feet long, fourteen wide, and fourteen high, without any pit in the 
middle. Having thus stated the dimensions of the houses, I shall now give the 
results of the investigation and calculations made relative to fuel, attention, &c. &c. 
The coals for both houses were measured before being placed for use ; and after the 
consumption of the night's fuel, the result was as follows : — The iron-roofed with 
18° of frost, required the consumption of nearly six bushels of coals, and unremitting 
attention during the night, or until three o'clock in the morning ; while the house 
with the wooden -roof consumed scarcely three bushels of fuel in order to keep it at 
the same degree of temperature with its iron rival ; and no attention was required 
after ten or eleven o'clock at night, when the fires were made up and left. More- 
over, being determined to investigate thoroughly the merits of the two materials, I 
caused a house, constructed of wood, and also one of iron of precisely the same 
dimensions as regarded superficial feet of glass, to be perfectly repaired in the 
autumn of 1832; and on having them examined and repaired in the following 
season, I found that in the cost of repairing, the iron house cost nearly double the 
sum required to repair the wooden one. I do not mean to say that double the 
number of squares were absolutely broken, but including the broken and cracked 
squares there was more than double the number destroyed, and this I attributed to 
the expansion of the iron during the summer and its contraction in winter. 
From these calculations, it is quite evident that wood has the advantage over iron in 
four very essential points, viz., the saving of /we/, glass^ and lalmir^ and in the better 
growth of plants and fruits, as I have invariably found plants do not thrive so well 
or look so healthy in an iron as in a wood-roofed house. The non-conducting power 
of wood, and the electrical (nay I may say caloric) sensibilities of iron may be 
the cause of this difference. Iron is infinitely more liable than wood to the sudden 
and injurious variations of temperature from heat to cold. I have always found 
during my practice, that no matter how the iron hothouse might be situated, unless 
there was a slight shading on the houses during the hot days in the summer 
months, the leaves of the pines and other plants become very brown and frequently 
scalded ; but whenever these shadings are not used, I would strongly recommend 
that a large cistern or trough of water be placed about the houses in order to make 
up by the continued evaporations for the deficiency of moisture exhaled by the 
powerful action of the sun. Another important circumstance is worthy of the 
