THE GARDENER. 



21 



that it would be less expensive to change their position 

 than to persevere in the culture of them ; but in form- 

 ing a new garden, he should exercise sound judgment 

 in the selection of its locality. A low and sheltered 

 situation, but not so low as to be without sufficient 

 inclination for discharging superfluous moisture, with a 

 southern aspect, is evidently the most desirable posi- 

 tion, but it is a mistake to suppose that a low situation 

 is always the warmest ; no doubt the shelter which 

 it possesses from parching and destroying winds is 

 greatly in its favour, but as far as mere warmth of 

 temperature is concerned, such a situation does not 

 command it, because, since cold air is heavier than 

 warm, the atmosphere nearest the earth will be cooler 

 than that which is higher : if then a garden is in a flat 

 in a valley, the cold air rests there — besides, the cold 

 air which issues from the sides of the hills falls into 

 the valley, so that what are by a common error sup- 

 posed to be warm situations, are in spring and autumn 

 the coldest. It is for this reason that potatoes, kidney 

 beans, and dahlias, are destroyed by frosts in autumn 

 in the valley of the Thames, while the same vegetables 

 are uninjured on the low hills of Surrey and Middlesex 

 at the same season. 



Professor Lindley found that one thermometer stood 

 in the same night thirty degrees higher on a slight 

 eminence, than the other did in the vale, fr^m which 

 he infers, that a garden should be on a gentle slope ; 

 and he points out the advantage of having a stream of 

 running water at the bottom of the garden, to prevent 

 any injurious stagnation of air. Shelter from the north 

 and north-east wind, which are the most destructive, 

 must also be a principal object. If there be no natural 

 shelter from those points, very hardy evergreens should 

 be planted outside the walls or other fences, but not too 

 near ; while every tree or other obstacle to the free ac- 

 cess of the sun's rays on the south side should be, if pos- 



