PLANTING AN ABBOKKTUM. 323 



available to our present purpose. It is hardly needful 

 to point out the immense disruption of affinities these 

 cliasms must cause, both in the alliances and tlie in- 

 ternal structure of the ordei-s. Hence the arrangements 

 of an arboretum can be at best only fragmentary ; but 

 that is no reason for neglecting or disregarding the 

 materials for combination which are within our reach. 

 The fragments that remain to us arc capable of assum- 

 ing a highly scientific form, and so may be invested 

 with attractive interest. 



Now, in planting the trees and shrubs of an arbore- 

 tum we miglit begin at one end of tiie classification, 

 it matters not at whiclj, and taking the first genus that 

 came to liand, we might put down its species in a 

 straight or curved line, or double line, and we might 

 proceed to the other genera successively, till we had 

 gone through the whole series, as Ave would wind off 

 a thread from a reel. The lines might be drawn along 

 a border within the four sides of an inclosure, or might 

 occupy narrow parallel borders, separated by walks, 

 or might assume the form of a spiral, running from 

 the exterior of the space to its center, or reversely. 

 This is what we have already called the linear arrange- 

 ment; but though it has been adopted in some nurse- 

 ries for the sake of mercantile convenience, it is highly 

 objectionable in various points of view. Not to speak 

 at present of the sacrifice of beauty thus made, it is 

 evident that this collocation preserves only the affini- 

 ties which a genus or order bears to that immediately 

 preceding, and to the other immediately succeeding; 

 and these undoubtedly do not include the whole of its 

 relations. As already noticed, Dr. Lindley, in his 

 " Vegetable Kingdom," appends to each of his orders 



