100 



may be exposed to a temperature a few degrees less than 32*^, 

 for a short time, without material injury, it is found that when 

 kept long at a temperature but little above freezing the ripen- 

 ing process is checked to such a degree that it is never per- 

 fectly resumed, and experience has shown that a uniform tem- 

 perature of 40*^ is as low as is advisable. Fruit may safely 

 be exposed to still air of less than S2° when it would be in- 

 jured in currents of the same temperature. It seems hardly 

 necessary to say that a thermometer is indispensable in a fruit 

 room. 



6th. Light is one of the strongest stimulants of vegetable 

 growth. Its effect on the skin of the fruit continues, though 

 in a less degree, after it is separated from the tree, and there- 

 fore the fruit should be kept as far as possible in total dark- 

 ness. Not only does it keep better but the flavor and color, 

 both of summer and winter fruit, is better when the light is 

 excluded. 



7th. The atmosphere should be uniformly dry, for mois- 

 ture is not only necessary to fermentation, but by being con- 

 densed on the skin tends to decompose it, and thus render its 

 protection less effectual. This happens to a greater extent in 

 apples and pears with smooth and glossy skins, than on those 

 russetted varieties, whose thick, rough skins are less perfect 

 conductors of heat, which may account for the well-known 

 long-keeping of the russet apples. The moisture produced 

 by condensation from the air must not be confounded with 

 that caused by sweating when the fruit is placed in large 

 masses soon after being gathered. When the air becomes 

 colder than the fruit evaporation takes place, and the surface 

 becomes dry, and these alternations of moisture and dryness, 

 like those of heat and cold, are injurious. 



Having determined the conditions necessary to the ripening 

 and preservation of fruit, we have now to apply these prin^- 

 ciples to the construction of a fruit house. Passing by the 

 costly patented methods which have been tried and con- 

 demned, or which, though correct in principle, are not yet 



