Practical Directions 277 



Single specimens of tall trees standing aniidst a tribe of very 

 much smaller ones look extremely naked and do not blend 

 at all beautifully or softly with the rest. Nor would the 

 hardness of their appearance be mitigated for several years. 



No plant will ever answer the expectations of the cultivator 

 if its roots be buried too deeply at the time of planting or 

 afterwards. Such a practice would shut them out from air 

 and speedily tell upon the health, most probably killing the 

 plant ultimately. The crown of the root ought not to be 

 placed more than two or three inches below the surface of 

 the ground. As the soil settles the plant will then at length 

 have the collar or crown of its roots just level with the ground, 

 and this is the most natural and healthy condition. 



That plants in masses should not be placed in any kind of 

 rows, but be dotted about as irregularly as possible and at 

 various distances from each other and from the front or back 

 of the plantation, would seem quite a trite remark, were it 

 not a rule that is seldom observed in small gardens. Nothing 

 is more common than to see the plants put in either straight 

 lines or rows following the outline of the mass, at one meas- 

 ured distance apart, and with two plants of the same kind 

 occupying precisely the same position in the bed on opposite 

 sides of the garden, thus making the arrangement of a group 

 a system of pairs, rather than the most inartificial and broken 

 thing imaginable. Even in some great public and national 

 works the trees are planted in rows although the outlines of 

 the plantations in which they occur are decidedly irregular. 



All this, however, unless where studied lines or avenues 

 are contemplated, is far too artificial for the natural style 

 of gardening, which is essentially free, varied, and approxi- 

 mating to nature. And since no such things as lines of 

 plants, symmetrical correspondence of sorts in particular 

 parts, or anything approaching to regularity of distance 



