36 
IRON AND WOOD ROOFS FOR STOVES, &c. 
gardener's attention : — viz., that iron houses should be painted internally, either 
annually or biennially at the farthest, to prevent the drip from the corroded iron 
injuring the foliage of the plants ; for I have always found this ochreous or 
metallic deposit injurious to the leaves. Since these remarks were made, I have 
had sufficient proof of their correctness, and of my former comparisons, through 
leaving extended my observations still farther in the season of 1834 ; and these, as 
you will perceive, fully confirm the accuracy of my previous calculations. By 
working two houses at the same temperature, 55 to 60° Fahrenheit, the result was 
as follows : — The wooden-roofed one consumed only one bushel and a half every night, 
while the iron-roofed one burnt from two and three quarters to three bushels ; this last 
experiment was made two months later in the season than when my attention was 
directed to the subject before ; but then you will see they bear the same propor- 
tions as to fuel, &c. The dimensions of the houses were as follows :-— the wood- 
roofed house fifty feet long, fourteen wide, fourteen high ; the iron, sixty feet long, 
fifteen wide, twelve high ; the latter was a vinery, and had a pit in it for the 
culture of pines, which very much reduced the number of cubic feet of air to be 
heated compared with the wood-roofed one for the culture of peaches, which had no 
pit in the centre. 
Notwithstanding however, that, the whole of my observations and calcula- 
tions are unfavourable to iron roofs, yet I am willing to admit, that for light- 
ness and neatness of appearance in the structure, iron has and always will have 
the preference ; but still I am confident, that if proper attention were paid to 
the construction of hothouses, and to the materials used in their erection, the 
appearance of wooden roofs would not be altogether objectionable. Now, for the 
assistance of persons building houses for horticultural purposes, who may have had 
less practical experience than myself, I will here give a brief description of such 
materials and mode of construction as I think will combine the whole of the 
desired objects. The first thing to be attended to, is to give the roof a proper 
pitch or inclination, so as effectually to carry off the water, and to prevent drip in 
the house, which is highly injurious to all plants, particularly those grown in pots. 
Secondly, to form the roof in the following manner :— the rafters to be of wood, 
varying according to the length of the roof from six to eleven inches, the section of 
the rafter to be wedge-shaped, from three to four inches wide on the upper side 
where the lights rest, and about half an inch wide on the bottom or under side. 
The ends and sides of the lights to be made of wood, the top from five to six inches, 
the sides two inches and a half, and the bottom from six to seven inches wide ; and 
the sash bars, to prevent as much as possible the obstruction of the sun's rays, 
should be of copper, which will give the house a light and neat appearance, without 
subjecting the plants to the injurious extremes of temperature, heat and cold, as the 
small quantity of metal in the thin sash bars, which need not be more than half an 
inch wide and about the same depth, will cause but very little variation in the 
temperature by radiation, and suffer little from expansion and contraction ; neither 
