PLANTS, WITH GLASS ROOFS. 179 



vary from 3 per cent, the quantity contained in sea-water, which will 

 not freeze when it is above 28, to 35 per cent, the greatest amount of 

 common salt which water will hold in solution. With 4*3 per cent of 

 salt, water freezes at 27|; with 6'6 per cent of salt, at 25|; and with 

 ll'l per cent, at 21 J. The effect which would be produced on cast- 

 iron pipes and boilers by any of these quantities of salt, Mr. Hood 

 states, would not be of much importance. As salt does not evaporate, 

 when a sufficient quantity is once added for the purpose required, the 

 waste which takes place can be supplied by fresh water. (* Hood's 

 Treatise,' p. 167.) In severe weather it is much safer to empty the pipes 

 or put on a little fire. All boilers should have a waste-pipe at their 

 lowest part, to empty them thoroughly at pleasure. 



Retaining Heat by Coverings. Whatever mode of heating plant- 

 structures may be adopted, it should be constantly borne in mind that 

 it is incomparably better for the health of the plants to prevent heat 

 from escaping by non-conducting coverings during the night, than to 

 allow it to be continually given off into the atmosphere, and as con- 

 tinually supplied by fire-flues or hot-water pipes. Where coverings 

 cannot be applied, and a high temperature must be kept up, reserve 

 sources of heat, and abundant supplies of water to maintain atmo- 

 spheric moisture, are the only means by which the plants can be kept 

 healthy. " A weakly growth," Sir J. Paxton observed, " is the sure 

 consequence of a high temperature maintained by fire-heat, whatever 

 plan of artificial heating be adopted." He therefore recommends, in 

 all cases where practicable, the use of external coverings, by which, at 

 Chatsworth, a difference of from 10 to 15 is gained, and two-thirds 

 of the fuel that would otherwise be necessary are saved. 



Atmospher-ic Moisture. The necessity of proportioning moisture to 

 temperature, and the causes which render the climates of our plant- 

 structures unnaturally dry, have already been pointed out. To give 

 an idea of the quantity of moisture requisite for an atmosphere at a 

 high temperature, Mr. Kogers has shown that a vinery twenty-five feet 

 long by thirteen feet six inches wide in the roof, maintained at 65 

 when the outer air is 35, will condense on the glass, in twenty-four 

 hours, 35^ gallons of water. (* Gard. Mag.,' 1840, p. 282.) In de- 

 vising the best method of procuring a constant supply of moisture for 

 the air of a hothouse proportionable to the expenditure, Mr. Rogers 

 finds the end may be most effectually attained by placing cisterns on 

 the heating-pipes. As the temperature of the water in these cisterns 

 would vary with that of the pipes, the evaporation from tljem would be 

 greatest when the pipes were hottest; when the greatest degree of 

 artificial temperature was being obtained, and consequently the greatest 

 drain upon it by condensation. The cisterns may be made of zinc, with 

 their bottoms fitted to the curvature of the pipes, at least six inches 

 deep to the top of the pipes, and of the same length as the space 

 between the rings by which the pipes are joined. Where two pipes are 

 placed side by side on the same level, the form shown in fig. 148 may 

 be adopted, and a single pipe may have cisterns fitted to it in the same 

 manner, or it may be made to embrace the sides of the pipe and cover 



N 2 



