CLASSIFICATION. 367 



and size has its difficulties in the uncertainty and varia- 

 tion of these characters in the different soils. and climates 

 o our extended country, and a sub-division and group- 

 ing of fruits by their quality of excellence is not only un- 

 reliable, but is altogether arbitrary, and subject to the 

 greatest diversity of opinion arising from the various 

 tastes of different individuals. We must look to some 

 marked and reliable characters that are always present, 

 easily recognized, and permanent or fixed. Among these 

 shape or figure stands pre-eminent, notwithstanding the 

 acknowledged fact that some varieties are almost protean. 

 The shape of the general outline appears to be the best char- 

 acter for the broad divisions of a classification. A sub- 

 division may again be made, which is to be based upon 

 the regularity or irregularity of the shape. 



The next character, and one of considerable value, is 

 that dependent upon flavor in its broadest characters of 

 sweet and sour, which, though sometimes giving rise to 

 a puzzling question, is, in most varieties, sufficiently 

 marked to constitute the basis of a minor sub-division. 

 Color, which is notoriously the poorest character and least 

 esteemed by botanists in their descriptions, on account of 

 its liability to variation, is, however, of sufficient import- 

 ance in pomology to take a high rank and to appear very 

 prominently in fruit nomenclature. Still it should be re- 

 served for the lowest sub-divisions of a classification. 



Among our American writers, who deservedly stand 

 prominent as pomologists, the most satisfactory attempt 

 at classification is found in the little work prepared by J. 

 J. Thomas. No one who has realized the advantages to 

 be derived from the simple and clear sub-divisions made 



