2150 ARBORETUM AND FRUTICETUM. PART 111. 



has been, first, to study the subject historically, that is, to ascertain what has 

 been said of it in books ; and, next, to study it practically, that is to compare 

 the information and the plates given in books, with living plants. After 

 perusing all the works we could procure on the subject, including Lambert's 

 Genus Pinus, 2d edit., 2 vols, 8vo, and the third volume of that work (which, 

 though only three or four copies have yet, August, 1837, been published, 

 we have been very kindly favoured with the loan of by His Grace the Duke 

 of Bedford), we took the first two volumes of Lambert's work, and that 

 volume of Michaux's North American Sylva which contains the ^Ibietinae of 

 North America, in our hands, and visited Loddiges's arboretum, the Horti- 

 cultural Society's Garden, Kew, Syon House, Dropmore, Whitton, Pains 

 Hill, Mill Hill, White Knights, and the principal nurseries; and, from the study 

 of the plants in these places, in connexion with the descriptions and plates 

 in the books we have mentioned, we have arrived at the general conclusions 

 which we shall now shortly lay before the reader, as preliminary to giving each 

 genus, and its species and varieties, in detail. 



In every arrangement of species and varieties, it appears to us that there 

 ought to be two objects in view. First, to throw all the kinds into groups 

 capable of being more or less distinctly defined j or, at all events, of being repre- 

 sented by one species as a type; such, for example, as the group Sylvestres, 

 of the section Binae, which consists of species all more or less resembling the 

 Scotch pine in foliage and in cones. The use of these groups is, to render 

 the whole mass easily comprehended by any person who knows only a few of the 

 species ; and, secondly, when separating these groups into species and varieties, 

 to give as prominent a place to all varieties and subvarieties that are truly dis- 

 tinct, as if they were species. Besides the argument which we have advanced in 

 favour of throwing the kinds into groups, there are the important ones men- 

 tioned in Part II. of this work (p. 216.) ; viz. those of assisting a collector of 

 trees to make a judicious selection, and of preventing a beginner in botanical 

 studies from puzzling himself in finding out specific distinctions where none 

 really exist. The reason why we wish to keep every variety and subvariety 

 as distinct as possible is, that, in the practice of arboriculture, whether for 

 useful or ornamental purposes, a variety is often of as much importance as a 

 species, and sometimes, indeed, more so : for example, in P. sylvestris, the 

 Highland variety is known and acknowledged to produce timber of a superior 

 quality to the common kind ; and, in point of ornament, for situations where 

 the common kind of Scotch pine is too large, the species may be repre- 

 sented by P. (s.) p. Miighus nana, which forms a beautiful little bush. 



In studying the y/bietinae from living trees, the terminal buds, the number 

 of leaves in a bundle or sheath, and their position on the branch, the sheaths 

 being persistent or deciduous, and the form of the cones, and the character 

 of their scales, are the principal points by which, we think, one species or group 

 of kinds can be distinguished from another. Thus, in Pinus, all the varieties of 

 P. sylvestris have short-pointed resinous buds, differing less in this respect 

 than they do in their cones, or in the length of their foliage. P. Laricio 

 (which we consider as including a number of European and some Asiatic 

 kinds, generally ranked as species, such as P. taurica, P. romana, P. cala- 

 brica, P. caramanica, &c.) is distinguished by its long, sharply pointed, con- 

 cave-sided, resinous buds; and P. Pinaster and P. Pfnea, by their short, 

 blunt, imbricated buds, which are never covered with resin. The buds of 

 P. TVda (which we consider to be the centre of a group of varieties generally 

 treated as species, under the names of P. rigida, P. variabilis, P. serotina, &c.) 

 are very small and resinous, and they are more numerous on the shoots than in 

 any other species, either European or American. All the kinds belonging to 

 P. TVda have also the peculiar property of sending out numerous small abor- 

 tive shoots from the dormant 01 adventitious buds in their trunks and larger 

 branches, by which the trees may be known at a glance, even at a distance. 

 All the tender kinds (such as P. longifolia, and its allied sorts, P. leiophylla, 

 P. canariensis, &c.) have small obscure bails ; and so on. The scales of 



