192 REVIEWS. 



tlie least difficulty ; but some (among wliicli are the Clalrgeau 

 and the Bosc) obstinately refuse to unite with the Quince 

 stock. He naturally discredits the assertion made by Cabanis 

 and by Downing (cited by Darwin), that when certain pears 

 are grafted on the Quince, their seeds produce trees of types 

 different from those which they do when they are raised upon 

 a Pear stock. Decaisne found, as already stated, that Pear- 

 seeds produce indifferently new varieties in any case ; that 

 these varieties are not at all fixed into races. He regards as 

 wholly unproven all the assertions that the fruit is ameliorated 

 or in any degree altered by grafting upon a Quince or any 

 other stock. He records a very exceptional instance in which 

 the antipathy of the Pear to the Apple as a stock was so far 

 overcome that the graft survived at least six years, but with- 

 out vigor, and bore fruit ; still this antipathy confirms the 

 generic difference between Pyrus and Mains. 



We must pass over the sections on the diseases of the Pear, 

 and the parasitic plants and insects hurtful to it ; while as to 

 that on the classification of the Pears of cultivation, we may 

 mention merely the conclusion, which is, that a natural classi- 

 fication of Pears is thus far an impossibility ; and that in 

 practice nothing better can be done than to follow the exam- 

 ple of the older pomologists, who arranged them according to 

 the period of ripening. A general list of the adopted names 

 of the published varieties of cultivated Pears, alphabetically 

 arranged, fills four pages of the volume. A list of their syno- 

 nyms, in which each is referred to the adopted name, fills over 

 12 pages ! Then follows a list of pears classed according to 

 the period of maturing, and in which the best varieties are 

 designated. 



Finally comes a botanical monograph of tlie genus Pyrus, 

 with a full generic character, and descriptions and figures of 

 the traces, as he would term them, considering as he does all 

 known forms of the restricted genus as a single and very 

 polymorphous species. 



The six races are : 1. The Celtic, Proles Armoricana, of 

 three quasi-species, P. corclata, Boissieriana, and longij^es. 

 2. The Germanic, Proles Germanica^ or Pyrus communis^ 



